Drinking Water and Health,: Volume 6 (1986)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

4 Neurotoxic Effects Interest in the toxicology of the nervous system has grown rapidly in recent years not only because of heightened public concern about the impact of toxic substances on human health but also because the nervous system has been shown to be especially vulnerable to chemical insult (Anger and Johnson, 1985~. The potential effects of such toxicity are quite varied, reflecting the large role played by the nervous system in regulating vital body functions and in perceiving, assessing, and responding to the external environment. Interference with these processes by overexposure to chemical and certain physical agents can produce a variety of abnormal neurobehavioral responses, some of which may be life threatening, others resulting in short-term neurobehavioral changes that disappear without a trace, and still others causing permanent or even progressive neurological or psychiatric disorders. Of special concern is the possibility that envi- ronmental exposure to certain neurotoxic agents, such as lead, may ir- reversibly compromise the normal development and capacity of the human brain. Although there are no reliable estimates of the number of persons who develop neurotoxic disorders in adult life, or of the impact of overexposure to chemical agents on the developing or aging nervous system, more than 850 chemicals are known to produce neurobehavioral disorders in humans or in animals (Anger and Johnson, 19851. The National Institute for Oc- cupational Safety and Health (NIOSH) considers neurotoxic and psycho- logical conditions 2 of the 10 leading work-related disorders, and the American Conference of Governmental Industrial Hygienists (ACGIH) has identified neurotoxicity as a basis for recommending threshold limit 105

i06 DRINKING WATER AND HEALTH values (TLVs) for 30% of the chemicals most frequently encountered in industry for which toxicological symptoms of occupational disease have been documented (Anger, 19841. Of the 197 industrial chemicals found in a 1974 NIOSH survey as having been implicated in exposure of more than 1 million people (NIOSH, 1977), 65 are neurotoxic at some exposure concentration. The magnitude of the worldwide occurrence of neurotox- icity in humans is also demonstrated by the recent World Health Orga- nization estimate that 375,000 cases of pesticide intoxication occur annually (Almeida, 1984~. Most pesticides exhibit some neurotoxic effects. Neurotoxicology is concerned not only with the actions of chemical agents but also with biological and certain physical (e.g., radiation) agents that produce adverse effects on the developing, mature, and aging nervous system (including special sense organs as well as neuroendocrine and neuromuscular systems) and on the behavior of humans and other animals. Some neuroactive agents produce rapidly reversible changes, other com- pounds cause permanent damage, and a few may induce progressive and terminal neural deterioration. Toxic disorders of the nervous system follow overexposure to abused substances (e.g., ethanol, inhalants, narcotics), therapeutic drugs, toxic products or components of living organisms (e.g., bacteria, fungi, plants, animals), chemicals designed to affect organisms unwanted by humans (e.g., fungicides, herbicides, insecticides), industrial chemicals, chemical warfare agents, additives and natural components of foods, and certain other types of chemicals encountered in the environment (Spencer and Schaumburg, 19801. The mechanisms of action of chemical agents that lead to nervous system dysfunction are, in general, poorly understood. Some agents act directly on the nervous tissue; others induce neurological or behavioral dysfunction indirectly, for example, by induc- ing changes in electrolyte balance, in cerebral blood flow, in metabolism of glucose, or in levels of critical intermediary metabolites. Such patho- physiological changes are often expressed clinically as neurobehavioral disorders because of the peculiar sensitivity of neural tissue to disruption of body homeostasis. Principles of neurotoxic response can be founded on the biology of the nervous system during development, at maturity, and during senescence. Experimental studies and clinical experience have demonstrated that the response of the developing nervous system to chemical insult may be either quantitatively or qualitatively different from the response of the mature system (Wilson and Fraser, 1977a, pp. 50-541. Some agents that induce developmental abnormalities of neural structure (neuroteratogens) or function (behavioral teratogens) may produce no or different abnor- malities in the adult (Vorhees and Butcher, 19821. Very few studies have been conducted to examine the susceptibilities of aged subjects to toxic substances. Thus, our understanding of the biological processes that might

Neurotoxic Effects 107 underlie differential responses in that population is limited. However, it seems reasonable to conjecture that deficits associated with old age are aggravated by those chemical agents that further impair neural centers and functions already compromised by the aging process. The House Select Committee on Aging has identified neurotoxicology as an important area of research likely to have a strong impact on the prevention and treatment of neurological disorders associated with old age (OTA, 19841. In the following paragraphs, the committee has provided a rationale for using neurotoxicity as a basis for determining acceptable levels of exposure to chemicals in community drinking water. The discussion includes a review of the structure and function of the nervous system, an outline of the significant forms and end points of neurotoxicity or end points that may serve as a basis for regulation, and the identification of susceptibilities of special populations. Later in the chapter, the types of information and research that can contribute to a risk assessment based on neurotoxicity are described, and issues relating to the use of neurotoxicity as a basis for risk assessment are discussed. THE HUMAN NERVOUS SYSTEM IN HEALTH AND DISEASE Gross Structure and Function By exploring different aspects of the vast body of knowledge on the human nervous system, one can attempt to learn how neural elements react to xenobiotic substances. Although the acquisition of such an un- derstanding begins with a study of the structural and functional organi- zation of the nervous system, the broad range of these neurotoxic effects is usually recognized first by the appearance of clinical symptoms or deficits in behavioral function. At maturity, the nervous system is separated anatomically into central and peripheral divisions. The peripheral nervous system (PNS) is com- posed of nerve cells (neurons) and their processes (axons), which conduct information to muscles and between muscles, glands, sense organs, and the spinal cord or brain. PNS axons are ensheathed by Schwann cells to form nerve fibers, which run together in bundles in peripheral and cranial nerves (I and III-XII) (Thomas and Ochoa, 19841. These nerves contain afferent (sensory) and efferent (motor) fibers, both of which are repre- sented in the somatic (voluntary) and visceral (autonomic) components of the nervous system. Somatic afferent fibers conduct information from special sense organs and sensory receptors in skin and muscles to the brain, whereas visceral afferent fibers convey impulses from the gut, heart and vessels, glands, and various organs. Somatic efferents innervate striated muscle, whereas visceral efferents supply smooth muscles of blood ves-

i08 DRINKING WATER AND H"LTH sets, glands, heart, and gut. In toxic states involving the PNS, degeneration of sensory and motor fibers leads to peripheral neuropathies associated with sensory loss (e.g., decreased sensitivity to vibration, touch, or phys- ical orientation) and motor (muscle) weakness. Dysfunction or breakdown of autonomic fibers may precipitate abnormal sweating, cardiovascular changes, or dysfunction of the gastrointestinal tract, the urinary tract, the genitals, or other organs or systems (McLeod, 19831. Manifestations of toxic disorders of the central nervous system (CNS) depend largely on the site, nature, and extent of the induced functional or structural change. The CNS consists of those parts of the nervous system contained within the skull and vertebral column. The spinal cord receives information via afferent fibers from PNS sensory receptors in the skin, voluntary muscles, tendons, blood vessels, and glands. It transmits signals for motor function through efferent fibers and communicates information via specific pathways, which include coordination centers within the brain. The brain is responsible for initiating, receiving, and integrating signals needed to maintain internal homeostasis, cognition, awareness, memory, language, personality, sexual behavior, sleep and wakefulness, movement and locomotion, sensation, vision, audition, balance, taste, olfaction, and many other body functions (Kandel and Schwartz, 19811. The brainstem, consisting of the midbrain, pans, and medulla oblongata, receives and processes information from skin, muscles, and special sense organs (e.g., inner ear), and in turn controls motor and certain autonomic functions relayed to the periphery by way of the cranial nerves. The cerebellum and basal ganglia are required for the modulation, regulation, and coordination of muscle tone and activity. The diencephalon, including the thalamus and hypothalamus, is a relay zone for transmitting information about sensation and movement. It also contains important control mechanisms for maintaining the internal homeostasis of the body. The hypothalamus functions as the primary control center for the visceral organs and inte- grates the activity of the neuroendocrine and other systems. The cerebral hemispheres, capped by the cerebral cortex, are concerned with perceptual, cognitive, motor, sensory, visual, and other functions. The optic (cranial II) nerves and their radiations conduct visual information from the retina through the thalamus to the occipital cortex. The cerebral cortex is divided into areas responsible for sensory recep- tion, motor control, and integration or association. Visual input is received in the posterior (occipital) cortex, and auditory input is directed toward the temporal lobes. Tactile sensation is perceived in the parietal cortex. The frontal cortex is concerned with the control of movement and is associated with the parietal cortex in the unilateral (left or right) regulation of speech. The remainder of the cortex is devoted to integrative and associative functions. Beneath the cerebral cortex lies a number of other

Neurotoxic Effects 109 structural areas, most with relatively clear functions. The basal ganglia, located at the base of the cortex, are important in the control of automatic and coordinated movement, as is the cerebellum. The limbic system con- sists of portions of the cortex and lower parts of the brain, which are involved in the control of eating, drinking, aggression, general arousal, motivation, and emotion. The primary or initial sensory areas of the brain are characterized by a receptor-topical organization; that is, the sensory receptor is represented point by point. Areas adjacent to each initial reception area in the cortex provide for an integration of the simple sensory input into successively more complex sensory perceptual organizations. One of the most thor- oughly explored of these areas is the visual system, whose initial reception area lies in the lateral geniculate ganglia of the thalamus, where a point- to-point map of the retina is organized. The somatotopic arrangement in a given portion of the retina is further organized by the orientation or angularity of the cell layers in the lateral geniculate ganglia and visual cortex as they appear in the retina. Specific neurons are responsive to edges of light and dark, angles, and corners—all of a particular size or strength. Such an organization appears well suited to the identification of complex visual stimuli (Kandel and Schwartz, 19811. The motor section in the frontal cortex controls movement throughout the body: the foot and leg are controlled at the top of the motor strip; the hand, at the top of the lateral aspect; and the face, jaw, and mouth, in most of the lateral area. An entirely separate section of the frontal cortex is devoted to the precise control of eye movements. By stimulating specific neurons in the motor area, individual muscles can be activated, whereas stimulation of several neurons can produce movements of various body parts. The motor area is generally divided into the pyramidal and extra- pyramidal systems. The pyramidal tracts descend from the frontal cortex to the spinal cord, where they regulate the activity of nerve cells controlling the skeletal muscles. The extrapyramidal pathways exert control over more general responses, such as postural adjustments, and do so through con- nection with other subcortical CNS structures, including the basal ganglia and cerebellum. The largest and least well understood area of the cortex involves as- sociative responses. The frontal association areas of the cortex govern delayed response or delayed-response alteration performance, mediated by memory. Disorders of the frontal lobe are often associated with eu- phoria, apathy, indifference, impaired judgment, recent memory, and immediate recall. When this area is removed, there are alterations of discrimination ability in a range of senses, especially when visual searching is involved. Hyperactivity, increased social aggressiveness, and inability to tolerate frustration generally occur when frontal lobe tissue is destroyed.

DRINKING WATER AND HEALTH A different set of functions is mediated by posterior cortical association areas, which govern visual-pattern discrimination and some aspects of color vision, but not visual acuity and object recognition. Learning ability is also mediated by this cortical area. The parietal association area is concerned with perception of body image and spatial relations as well as with learning involving somatesthesia. Disorders of language and inability to perceive objects occur in patients with a damaged parietal lobe (right or left). The two hemispheres of the brain commonly have structural differences and are functionally distinct with regard to higher functions. In a greatly oversimplified but didactically useful way, the human brain consists of two halves: a left hemisphere that tends to be an intellectual, rational, verbal, and analytical thinker, and a right hemisphere that tends to be a perceiver and an emotional, nonverbal, and intuitive thinker. In principle, each hemisphere is capable of independent function, but integration of function is normally maintained by an extensive network of interconnec- tions known as the corpus callosum. Hemispheric dominance develops gradually, as evidenced by the ability of young children to develop vir- tually normal speech and language capabilities after complete removal of the left hemisphere. In adults, extensive damage to the dominant hemi- sphere results in severe functional deficits in speech functions, and re- covery is slow or absent. Moreover, there are differences in the rates at which lateralization of higher functions develops in boys and girls. Hence, in preadolescents there are sex differences in vulnerability to left hemi- sphere damage. Speech becomes fixed at an earlier age in boys, who are therefore more likely to suffer more severe and enduring language deficits after cortical injury in childhood. Cellular Structure and Function The fundamental neurocellular components are neurons and glial cells, which are associated with blood vessels and other specialized epithelial and connective tissue cells (Jacobson, 19784. Neurons have relatively small cell bodies (perikarya) and multiple, relatively short processes (den- drites), which receive information transmitted from axons of other nerve cells. In addition to dendritic processes, neurons are characteristically equipped with a single, elongated axon that conducts electrical signals between neurons, or to muscles, skin, and glands. Signals are transmitted along axons by a longitudinal spread of changes in membrane potential— a difference in electrical potential that exists between the intracellular (axoplasmic) and extracellular environments. In the resting state, mem- brane potential is maintained by a high intracellular concentration of po- tassium ions and a high extracellular concentration of chloride and sodium

Neurotoxic Effects ~ ~ ~ ions. This balance is maintained by an energy-requiring ion pump located in the cell membrane; failure of this pump in certain toxic states leads to a disturbance of transmission and intracellular edema the accumulation of abnormal amounts of fluid inside the cell. Transmission of an electrical impulse along an axon is initiated by reversing the membrane potential locally; this is achieved by a rapid influx of sodium ions via membrane channels, cessation of sodium-ion flow, and an increase in potassium-ion conductance causing the potassium ions to move out of the cell along their concentration gradient. The resulting localized change in membrane po- tential the action potential—moves along axons by reproducing these electrical events in adjacent areas of excitable membrane. Impulse prop- agation is accelerated in myelinated nerve fibers by the concentration of electrical events on the naked axon at sites between adjacent myelinating cells (nodes of Ranvier). Toxic substances that interfere with the function of ion channels in the excitable membrane or impair the function of Ranvier nodes produce transient alterations of nerve conduction that precipitate characteristic sensory or motor phenomena (Narahashi, 19841. Axons terminate at synapses where chemically encoded information is conveyed to other neurons or effecter organs. Proven chemical neuro- transmitters (up to 50 are suspected) include acetylcholine, nonepine- phrine, epinephrine, and dopamine; putative transmitters include glycine, glutamic acid, ~y-aminobutyric acid, serotonin (5-hydroxytryptamine), his- tamine, substance P. enkephalin, and many other neuroactive peptides (Cooper et al., 1982~. Enzymes responsible for the synthesis of transmitters are manufactured in neuronal perikarya. Subsequently, with membranous and cytoskeletal elements, they move toward the nerve terminal by energy- requiring axonal transport (Brady, 1984~. Retrograde transport mecha- nisms return materials to the cell body for reprocessing (Kristensson, 1984~. Toxic substances may be specific for one or more transmitter systems, acting by changing the rate of synthesis, synaptic release, reup- take, or degradation or by interfering with the interaction of the chemical signal with the receptor surface on the dendrite or other cellular membrane (muscle or gland). Such changes are usually short lasting and reversible. By contrast, chemical agents that induce loss of CNS neurons produce irreversible damage. Degeneration of CNS axons, initiated by blockade of axonal transport or by other methods, is also commonly irreversible, except for small-diameter unmyelinated monoaminergic axons that regen- erate and elongate after injury. Glial cells in the CNS include astrocytes, oligodendrocytes, and mi- croglia (Varon and Somjen, 19791. Astrocytes are divisible into proto- plasmic and fibrous forms. They are closely associated with neurons and blood vessels, play a nutritive role in maintaining neurons and other cells, and regulate the composition of the extracellular fluid. Astrocytes have

~ 12 DRINKING WATER AND HEALTH also been implicated in the metabolic activation of methylphenyltetrahy- dropyridine (MPTP), an agent that produces parkinsonism in humans and animals. In addition, they provide an outer cellular border around the brain and spinal cord. These two organs are surrounded by a delicate pial connective tissue and are suspended in cerebrospinal fluid (CSF), which circulates between the ventricular system of the brain and spinal cord and spaces beneath the arachnoid and aural connective tissues that encase and protect the CNS. Interruption of the normal flow of CSF during devel- opment or adult life leads to hydrocephalus, a condition resulting from the accumulation of fluid in enlarged brain ventricles or between the brain and the overlying aura (Weller et al., 19831. The oligodendrocyte, another type of glial cell, elaborates short lengths of myelin around clusters of adjacent CNS axons. Myelination begins in the fourth month of fetal life and continues until the first year of postnatal life (Pansky and Allen, 19801. Toxic substances that induce changes in oligodendrocytes or myelin in postnatal or adult states cause reversible myelin vacuolation, demye- lination, and remyelination pathological events that interrupt nerve-im- pulse transmission without disrupting the structural continuity of axons (Rasminsky, 19801. CNS injuries resulting in a loss of tissue are associated with proliferation of astrocytes, which form a glial scar. In the PNS, Schwann cells envelop many small axons to form un- myelinated fibers, or associate with and elaborate lengths of myelin (in- ternodes) around single axons. PNS myelination occurs over a time course similar to that in the CNS. Toxic conditions resulting in damage to mature PNS axons are usually reversible, since axonal regeneration, reconnection, and reactivation of denervated end organs usually follow injury. Motor denervation leads to neurogenic atrophy of voluntary muscle. Damage to Schwann cells or myelin causes localized demyelination and then re- myelination pathological changes associated with blockade and resto- ration of nerve-impulse activity, respectively (Rasminsky, 19801. Exten- sive demyelination may be accompanied by axonal loss and poor functional recovery. The adult nervous system is provided with specialized regulatory in- terfaces, known as blood-brain, blood-nerve, blood-CSF, and blood- retinal barriers, among others. Such permeability barriers are formed by tight iunctions between adjacent endothelial cells lining neural capillaries and their corresponding basal laminae (Jacobs, 19801. The barrier func- tions as a differential filter that controls the passage of various substances (including certain blood-borne toxicants) from the blood to the interstitial fluid. Certain areas of the brain and PNS are devoid of blood-nerve bar- riers, including the area postrema, hypophysis, pineal body, hypothalamic regions, subfornical organ, supraoptic crest, choroid plexus, dorsal root, and autonomic ganglia. Such regions are directly exposed to toxic and

Neurotoxic Effects 113 other substances circulating in the blood. Experimental studies demon- strate that large doses of several glutamate analogs (excitotoxicants) induce neuronal degeneration in areas of the CNS lacking a blood-brain barrier (Olney, 1980) and that other agents (e.g., doxorubicin and large concen- trations of pyridoxine) alter the integrity or cause degeneration of sensory neurons in dorsal root ganglia (Cho et al., 1980~. The CNS and much of the PNS are also protected from exogenous agents by connective tissue sheaths, but the terminal regions of efferent PNS axons in muscle and glands may be directly exposed to noxious agents present in tissue fluid. Such agents may be selectively or nonspecifically taken up by nerve terminals and carried by retrograde transport to remote target sites in the neuronal perikaryon (for agents such as ricin) or associated synapses (for substances such as tetanospasmin) (Price et al., 1975~. Normal Developmental Structure and Function Development of the human nervous system begins during the third week of embyrogenesis (see Chapter 2 on Developmental Toxicity) (Pansky and Allen, 19801. The CNS commences as an elongated neural plate of thick- ened ectoderm. The lateral edges of the plate elevate, abut, and from the fourth week onward, fuse to form a closed tubular structure (neural tube) with a long caudal portion (the future spinal cord) and a broader cephalic segment (the future brain). The latter displays three distinct dilatations corresponding to the fore-, mid-, and hindbrain vesicles. Differential growth and migration of cell populations then mold the vesicles into more complex forms. By the fifth fetal week, the forebrain consists of the diencephalon with primordial optic vesicles, pineal gland and posterior hypophysis, and two lateral expansions the primitive cerebral hemispheres. The hemi- spheres expand rapidly and cover the brainstem, forming various spe- cialized centers including the hippocampus. During late fetal life, the surface of the cerebral hemispheres grows so rapidly that convolutions (gyri) develop on its surface. The diencephalon also gives rise to the thalamus, which contains the dorsal nuclei important for the reception and transmission of auditory impulses, and ventral nuclei, which serve as relay stations for transmitting the impulses to higher cen- ters. These and other elements form the limbic system, which is concerned with the regulation of emotional activity, visceral and memory functions, and such other activities of the hypothalamus as the regulation of body temperature. The pituitary gland (hypophysis) develops from a downward extension of the diencephalon (becoming the posterior hypophysis or neu- rohypophysis) and an ectodermal outpocketing of the primitive oralLcavity (Rathke's pouch, becoming the anterior hypophysis, or adenohypophysis). Each optic vesicle forms an optic cup, which grows out to form the optic

~ ]4 DRINKING WATER AND HEALTH nerve and retina, and induces the surface ectoderm to invaginate and form the lens and cornea. Congenital abnormalities of the eye induced by chem- ical and other agents may lead to the development of a single eye (cyclops), no eye (anophthalmia), malformation of the iris, retinal dysplasia or her- niation, and many other conditions (Wilson and Fraser, 1977b, pp. 329- 3411. In the normal state, after the eye opens postnatally, visual impulses are conducted from the retina to the lateral geniculate body and then to the occipital cortex. The remaining portion of the brain develops from the specialization of plates of cells. In general, cells located in the basal plate form motor neurons whose axons supply cranial nerves, while the more dorsal, alar plates form the sensory relay nuclei of the brainstem (Pansky and Allen, 19801. The midbrain develops the motor nuclei of cranial nerves and relay nuclei for visual and auditory reflexes. The hindbrain comprises the met- encephalon, from which develops the pons and cerebellum, and the myel- encephalon, the forerunner of the medulla oblongata. The roof plate of the myelencephalon forms the choroid plexus—tuft-like imaginations that produce CSF. The cerebellum develops from the metencephalon, and by the twelfth fetal week, forerunners of the midline vermis and paired hemi- spheres are visible. Cells migrate to the surface to form the external granular layer, and at the sixth developmental month, cells migrate from this layer inward toward the differentiating Purkinje cells. The inner ear, which subserves both audition and equilibrium, develops after the third week from otic placodes, which are visible on each side of the hindbrain. During later development, each vesicle divides into a ventral portion that gives rise to the saccule and cochlear duct and a dorsal component that makes up the utricle, semicircular canals, and endolym- phatic duct. The organ of Corti, the true organ of hearing, is visible by the tenth week. At maturity, impulses are transmitted by the spiral ganglion to the brainstem via cranial nerve VIII. The semicircular canals appear at the sixth fetal week, and at maturity, impulses generated in the sensory cells of the cristae and maculae as a result of position changes of the body are carried to the brain by the vestibular fibers of cranial nerve VIII. Other special sense organs include receptors for taste, which convey information via sensory ganglia to the brainstem and cortex, and receptors for olfaction, which send impulses directly to the primary and secondary olfactory cor- tex. Development of the spinal cord is comparable to that of the brainstem. Basal plates on the ventral sides of the neural tube form motor neurons, notably anterior horn cells, while the dorsal, alar plates form sensory association areas. These receive the central axon processes of sensory neurons in dorsal root ganglia, which develop from neural-crest cells that

Neurotoxic Effects ~15 overlie the dorsal surface of the neural tube. The peripheral axons of the sensory neurons supply sensory receptors. The motor neurons of the autonomic division of the PNS develop be- tween the ventral and dorsal plates deep in the immature spinal cord. A sympathetic branch is formed in the thoracolumbar region, and a para- sympathetic component is found in the craniosacral region. The efferent portions of both parts are composed of a chain of two neurons, whereas afferent impulses from pressure and chemoreceptors in the aorta and ca- rotid arteries ascend via axons of visceral sensory neurons in dorsal root ganglia (Mayer, 19801. Some sympathetic neurons originate from the neural crest and migrate on each side of the cord to form a bilateral chain of segmentally arranged ganglia interconnected by nerve fibers; others form the celiac and mesenteric ganglia of the aortic branches and the organ plexuses found in the heart, lungs, and gastrointestinal tract. Preganglionic efferent axons of the sympathetic chain use acetylcholine as their neu- rotransmitter. Postganglionic efferent neurons, in conjunction with the adrenal medulla, release nonepinephrine directly or through the blood- stream to effecter sites in the ciliary body of the eye, the salivary glands, bronchi, heart and blood vessels, liver, gut, kidney, bladder, and external genitalia. Preganglionic neurons of the parasympathetic system, located in the midbrain, medulla oblongata, and sacral spinal cord, also use ace- tylcholine as a neurotransmitter, as do their postganglionic counterparts located close to their effecter sites in the organs listed above. Functionally, the two divisions of the autonomic nervous system tend to counteract each other (Mayer, 19801: the sympathoadrenal system, which can discharge as a unit during rage or fright, accelerates heart rate; increases blood pressure and glucose concentration; shifts blood from the spleen, skin, and gut to the skeletal muscles; and dilates the pupils and bronchioles. By contrast, the parasympathetic system tends to conserve energy by slowing the heart rate, lowering blood pressure, stimulating gastrointestinal movements and secretions (thereby aiding nutrient ab- sorption), protecting the retina from excessive light, and emptying the urinary bladder and rectum. Synchronized discharge of parasympathetic neurons does not occur normally, but is seen during poisoning with anti- cholinesterases, which inhibit the enzyme (cholinesterase) required to ter- minate the transmitter activity of acetylcholine (Mayer, 19801. Neuroteratology and Psychoteratology Experimental studies and clinical experience have demonstrated that the response of the developing nervous system to chemical substances may be either quantitatively or qualitatively different from the response

~ ~ 6 DRINKING WATER AND HEALTH of the mature system. Some agents that induce developmental abnormal- ities of neural structure (neuroteratogens) or function (behavioral terato- gens) may produce no abnormalities in the adult (Vorhees and Butcher, 19821. Classic examples are agents that interfere with cell proliferation: these agents can alter the number of nerve cells permanently, if exposure occurs during neuron production, but may be benign or produce different effects in the same tissue when rapid mitotic activity has ceased. For example, although the mitotic inhibitor colchicine blocks axon transport at any age, its disruption of mitosis is problematic only in the immature CNS (in the adult, colchicine produces axonal neuropathy). Finally, at different stages of development, the CNS may have varying sensitivity to agents, even though the same mechanisms may be active. Classic teratology studies have demonstrated that the degree of respon- siveness of the conceptus to the induction of congenital malformations by teratogens depends on, and varies with, gestational age (Wilson and Fraser, 1977a, pp. 50-54~. As far as teratological susceptibility is concerned, the time from conception to birth is roughly divisible into three periods: before germ-layer formation, the period of embyro development, and the fetal stage. The first period, from fertilization to implantation, lasts approxi- mately 6.5 days in humans and is generally resistant to the induction of congenital malformations. The embyronic period begins after implantation and lasts to the end of the second gestational month. During this period, differentiation, mobilization, and organization of cells and tissue groups take place to form individual organ systems, and toxic disturbances of normal development while these processes are occurring may cause gross structural abnormalities (malformations). Since the genesis of different organs occurs at different stages of development, the type of malformation is largely dependent on the susceptibility of the individual organs at the time of the insult. Teratogenic agents causing interruption of neural-tube closure lead to birth defects such as spine bifida and anencephaly. After the neural tube closes and early development of the brain is under way, toxic insults may result in grossly abnormal brain morphology. During this stage, developmental abnormalities have been produced in experi- mental animals exposed to vitamin A (Wilson and Fraser, 1977a, p. 270) and trypan blue (Wilson and Fraser, 1978, pp. 116-117), the former producing anencephaly, anophthalmia, spine bifida, and other defects. Teratogenic susceptibility decreases in the third (fetal) stage, since cel- lular proliferation and differentiation is less marked and organogenesis may be complete (Wilson and Fraser, 1977a, pp. 50-541. However, some tissues, such as the cerebrum and cerebellum, continue active differen- tiation and thus remain susceptible to the action of certain chemical agents until term or even into the postnatal period. Experimental animal studies have demonstrated that neuroteratogenic changes can be induced in the

Neurotoxic Effects 117 fetal stage by transplacental exposure to methylazoxymethanol (MAM), 5-azac.ytidine, ethylnitrosourea, and x-irradiation. MAM also induces changes in cerebellar development when administered to laboratory ani- mals postnatally. Similarly, agents that inhibit myelination can have pro- found effects on oligodendrocytes when administered prenatally or postnatally (Le Quesne, 1980~. Finally, the immaturity of the blood-brain, blood- retinal, and perhaps other barriers may lead to special susceptibilities in developing states. The high incidence of lead encephalopathy in the young has been attributed to increased access of the circulating toxicant to brain tissue. Toxic agents act with some selectivity on the developing organism, inducing a characteristic pattern of teratogenic change within the frame- work of a specific developmental stage. However, an agent may produce different responses at identical levels of exposures occurring at different developmental stages. Differential reactions across species have also been recognized. Dose is another important factor: large doses can produce fetal death; smaller doses, teratogenesis; and the smallest doses, no detectable gross malformation. The thalidomide tragedy demonstrated graphically that the dose required to induce phocomelia in humans (0.5 to 1.0 mg/kg/day) differed markedly from that required to produce comparable effects in common laboratory species (10 to 350 mg/kg/day) (Wilson and Fraser, 1977a, p. 3141. Nutritional status of the dam is also a critical consideration: both fasting and specific vitamin deficiency states can induce disorders such as hy- drocephalus and eye defects in developing animals. Additive effects have been observed in fasting animals exposed to teratogenic substances such as trypan blue (Wilson and Fraser, 1977a, p. 451~. Mineral deficiency (e.g., manganese and copper) in the dam reportedly can lead to congenital ataxia in the offspring (Shils and McCollum, 19431. However, maternal deficiency of certain other required substances, including vitamins C, D, and K, fatty acids, choline, biotin, and various amino acids, fails to produce congenital malformations of the nervous system. Other experimental factors positively correlated with fetal susceptibility to the teratogenic effects of chemical substances include maternal weight and, probably, age, litter size, implantation site, route of administration, and season. Ionizing radiation, hypoxia, and certain infections supplement the long list of factors associated with neuroteratogenesis in humans and animals. Other neuroteratogens include aLkylating agents, certain anti- metabolites and alkaloids, hypoglycemic agents, ethanol, salicylates, an- tibiotics, antihistamines, neuroleptics, anticholesterolemic drugs, steroids, sulfonamides, and organomercury (Wilson and Fraser, 1977a, pp. 317- 340~.

DRINKING WATER AND HEALTH Psychoteratology, or behavioral toxicology, is a more recent discipline concerned with the study of the more subtle effects of chemical substances that induce changes in the nervous system or behavior of the developing organism (Vorhees and Butcher, 19821. Many behavioral teratogens are simply lower doses of classical teratogens that at these levels of exposure fail to produce gross malformations in the nervous system. Some of these behavioral changes have been traced to abnormal development of neu- rocellular elements detectable with the light microscope. Others, the psy- choteratogens, induce experimental behavioral changes without presenting detectable structural damage to the CNS and, even at higher doses, fail to induce malformations. The mechanisms underlying these changes are unknown. The most important example of human psychoteratogenesis is the fetal alcohol syndrome; other fetal syndromes have been reported in association with maternal exposure to methylmercury, barbiturates, phenytoin (diphenylhydantoin), trimethadone, and primadone. Neurotoxic Responses after Birth and at Maturity The nervous system of humans and other animals is also susceptible to toxic perturbations after organogenesis is complete, although the sub- stances that induce these changes and the nature of the disorders are usually different from those that affect the developing organism. As before, factors such as species type, dose and duration of chemical exposure, and the cellular target of the toxic agent or its metabolites are critical in determining the nature and severity of the induced lesion. Different syndromes may appear in humans according to the rate and degree of intoxication; for example, acute overexposure to acrylamide causes a toxic encephalopathy with seizures, whereas prolonged, low-level intoxication produces a syn- drome of peripheral neuropathy. Evidence suggests that most direct-acting neurotoxicants induce struc- tural or functional changes, as a result of either reversible or irreversible binding of toxic substances to receptors or other vital macromolecules in nervous tissue. There appears to be no good reason why neurotoxicity will not follow normal sigmoidal dose-response relationships. However, the dose-response curve for neurotoxic effects is likely to be steep, since similar degrees of qualitatively identical neurobehavioral effects com- monly occur in the majority of people or animals exposed to the same dose of the neurotoxic agent. For example, most, if not all, patients undergoing a standard regimen of vincristine for the treatment of a ma- lignant tumor develop peripheral (toxic) neuropathy (Schaumburg et al., 19831. Universal responses of exposed human populations are quite dif- ferent from the experience with chemical carcinogens where, in most instances, only a small portion of a population is expected to develop

Neurotoxic Effects ~19 malignant growths. Whether these statements are true for low levels of exposure to neurotoxic agents is unknown, and much more research is required to gather this information. Differential responses may appear in individuals of different sex, e.g., males are more susceptible to the Lathyrus toxin (Gopalan, 1950), and of different age. Structure-activity relationships are clear for only a few classes of compounds, such as the anticholinesterases, organophosphates, and hydrocarbon solvents with a common y-diketone metabolite (Johnson, 1975a,b; NRC, 1982; O'Donoghue, 19851. For other compounds, pre- diction of neurotoxicity based on chemical structure alone is presently a hazardous venture. The neurotoxic potential of some substances can be altered as a function of metabolic state, such as the susceptibility of slow acetylators (who are homozygous for an autosomal recessive gene) to the antituberculosis drug isoniazid (Ells, 1984~. Between 50% and 70% of the North American and Central European populations are slow acetyla- tors. Fast acetylators (i.e., those who are either heterozygous or homo- zygous for a dominant gene) constitute approximately 90% of the Japanese population (Price Evans, 19781. The potency of other neurotoxic agents may be altered by concurrent exposure to another compound lacking the property. This principle is clearly illustrated by the ability of methyl ethyl ketone to accelerate the development and increase the severity of the peripheral neuropathy induced by repeated exposure to n-hexane or methyl n-butyl ketone (Altenkirch et al., 19821. Developing countries are concerned that environmental compounds with neurotoxic potential may be especially hazardous to people with neuro- logical susceptibility associated with malnutrition. The synergistic action of ethanol and thiamin deficiency exemplifies this concern. Finally, there are some agents (e.g., organoarsenicals) that do not obey classic dose- response laws and unpredictably induce rapid-onset neurological disorders in humans comparable to the Guillain-Barre syndrome. It seems likely that these will prove to be hypersensitivity reactions resulting in cell- mediated attacks on targeted nervous system components, notably myelin. Alteration of the immune system was also implicated in the 1981 Spanish toxic oil syndrome a multiphasic, multisystem vascular disorder affect- ing some 20,000 people exposed to an adulterated cooking oil. Some victims of the syndrome eventually developed a devastating neuromuscular disorder in which muscles, nerves, and skin were invaded by fibrotic tissue. Toxic chemicals are often grouped and studied on the basis of their commonplace occurrence, use, or physiochemical properties. For exam- ple, free reference is made to industrial, biological, and heavy metal neurotoxicants. This method of classification is a necessary but entirely misleading practice, since the neurotoxic properties of a substance are

|20 DRINKING WATER AND HEALTH TABLE 4-1 Classification of Neurotoxicants by Target Target Prenatal period Neural tube Developing nervous system Postnatal period Neurons Excitable membrane Neurotransmitter systems Cholinergic agonists Cholinergic antagonists Structural integrity Neuron Axon Neurotoxicant Excessive vitamin A intake Ethanol, methylazoxymethanol Channel agents, including pyrethroids Anticholinesterases Tricyclic antidepressants Dendrite Glial cells Myelinating cells and myelin Oligodendrocytes (CNS) Schwann cells (PNS) Astrocytes Ependymal cells Special sense organs Olfactory/gustatory Optical Otic Vestibular Muscles Striated muscles Cardiac muscles Neural vasculature Neuroendocrine system Hypothalamus/hypophysis Immune system, with secondary ef- fects on nervous system or mus- culature Various cell types Malignant transformation, with primary or secondary growths affecting nervous system Doxorubicin, mercury, trimethyltin Acrylamide, n-hexane, methyl n-butyl ketone Glutamate excitotoxins Isoniazid, triethyltin Lead, diphtheria toxin 6 - Amino nic otin amide Amoscoline Penicillamine, thiouracil Methanol, chloroquine Noise, toluene Hydroquinone Anticholinesterases, dimethyl sulfoxide Diphtheria toxin Cadmium Chlordecone Gold thiourea Alkyl nitrosoureas based on chemical structure and target site in the nervous system—not on the source of the agent, type of usage, or, for most elements, position in the periodic table. A more appropriate classification of chemical neu- rotoxicants is based on apparent target sites within the nervous system (Table 4-1), although one agent may have more than one site of action,

Neurotoxic Effects 121 especially when different doses are under consideration (Spencer and Schaumburg, 19841. This nosological approach, i.e., the classification by apparent target site within the nervous system, can yield useful information, a point well illustrated by a diverse collection of chemical substances that interferes with the opening and closing of sodium channels in excitable membranes, thereby producing a common initial clinical expression of rapid-onset circumoral and distal-extremity paresthesias. The agents in question- tetrodotoxin, batrachotoxin, ciguatoxin, dichlorodiphenyltrichloroethane (DDT), cyanopyrethroids, and scorpion toxin—originate from diverse sources and have disparate chemical structures, but because of their similar sites of action, they produce comparable effects. These so-called sodium- channel toxins, and those that perturb neurotransmitter function, usually produce rapidly reversible changes in nervous system function; although if intoxication is severe enough to impair vital centers, such as the brain- stem respiratory center, death may ensue. Since the amount and timing of neurotransmitter release at synapses are critical, toxicity is associated with agents that increase the amount or mimic the action of neurotransmitters (agonists) as well as those that impair their function (antagonists) (Spencer and Schaumburg, 1984~. This prin- ciple is well illustrated by responses of the cholinergic system: cholinergic toxicity follows overexposure to direct agonists (e.g., muscarine) or agents that inhibit enzyme inactivation of acetylcholine (anticholinesterases), whereas anticholinergic toxicity is caused by agents such as 5-bungaro- toxin, which block acetylcholine reception at the neuromuscular junction (Mayer, 19801. Similarly, in the catecholamine system, some antidepres- sants inhibit monoamine oxidase, thereby increasing the duration of action of synaptic norepinephrine and circulating epinephrine, whereas certain hypertensives, such as guanethidine, interfere and displace catecholamines from presynaptic terminals. Other agents such as lysergic acid (LSD) and trimethyltin interfere with the action of serotonin, picrotoxin with y-ami- nobutyric acid (GABA), ibotenic acid with glutamate, tetanospasmin with glycine, or antipsychotics or opiates with several transmitters concurrently (Damstra and Bondy, 1980~. The type of neurological disorder associated with these different conditions usually reflects closely the functions sub- served by the neurotransmitter systems that have been perturbed by the toxic substance or its metabolites. Most of the disorders are not known to be associated with detectable structural changes in the nervous system or elsewhere, and a large majority of them are rapidly reversible. Some, such as the tardive dyskinetic states associated with chronic use of phe- nothiazines and certain other neuroactive drugs, are usually attributed to increased sensitivity of receptor sites on synaptic membranes and are irreversible, or slowly reversible after cessation of exposure.

}22 DRINKING WATER AND H"LTH Toxic disorders associated with compounds that produce structural changes in neurons, with consequent degeneration of the nerve cell perikaryon or distal axon, often develop slowly and lead to long-lasting or permanent decrements in sensory, motor, and autonomic functions. Some agents, such as the methylated compounds of lead, arsenic, or mercury, readily penetrate brain tissue and produce rapid, widespread, and irreversible neuronal degeneration. Other inorganic compounds (e.g., salts of thallium, arsenic, lead, aluminum, and bismuth) also are capable of producing encephalopathies with permanent neurobehavioral defi- cits, although the precise pattern of damage in these conditions has not been adequately described. Delayed and then progressive deterioration of brain structure and function may occur in persons who recover from acute carbon monoxide intoxication (Ginsberg, 1980~. Loss of neurons may also be observed in dorsal root ganglia after exposure to mercury or doxorubicin, whereas adjacent motor neurons protected by the blood- nerve regulatory interface in the spinal cord are usually spared. Retinal ganglion cells undergo degeneration in methanol intoxication, and other agents (e.g., quinine and lead) induce degeneration of photoreceptors (Merigan and Weiss, 1980~. Aminoglycoside antibiotics (e.g., strep- tomycin) cause loss of the receptor cells of the organ of Corti, and certain solvents (e.g., toluene, styrene, and xylenes) recently have been implicated as ototoxins in experimental animals (Lane and Routledge, 1983; Rebert et al., 19821. A particularly common type of neuronal change induced by toxic chemicals is central-peripheral distal axonopathy, i.e., degeneration of long and large-diameter axons in peripheral nerves, spinal cord, medulla oblongata, and cerebellum. Some agents (e.g., thallium and arsenic salts and organophosphates) can precipitate this type of change in hu- mans after a single intoxicating event; however, most other recognized neurotoxic substances of this class, such as n-hexane (Spencer et al., 1980a), acrylamide (see Chapter 9), and carbon disulfide (Seppalainen and Haltia, 1980), require prolonged intermittent or continuous expo- sure. Similar disorders are seen in common nutritional deficiencies (e.g., deficiencies of the B vitamins) and metabolic disorders (e.g., diabetes mellitus) in humans. Degeneration of axons begins distally and proceeds along affected tracts in a retrograde manner, a pattern referred to as dying-back. Stocking-and-glove sensory-motor neuro- pathy is the common clinical result of this pattern of damage, and considerable recovery usually follows cessation of exposure. However, if substantial degenerative changes have occurred in vulnerable as- cending and descending spinal-cord tracts, people who recover from PNS changes may be left with residual sensory loss, ataxia, or spasticity associated with permanent CNS deficits. A small number of agents

Neurotoxic Effects 123 (e.g., clioquinol) produce comparable disorders selectively in CNS tracts (e.g., central distal axonopathy) (Thomas et al., 1984~. Damage to myelinating cells may lead to life-threatening encepha- lopathy (e.g., from exposure to hexachlorophene) or neuropathy (e.g., from exposure to diphtheria toxin) (Cammer, 1980~. If the damage is restricted to myelin, and does not involve neighboring cells, a reason- able degree of recovery usually accompanies remyelination. Agents that damage myelin are legion: some appear to precipitate CNS and PNS demyelination without damaging the myelin-producing glial cells, whereas others (e.g., diphtheria toxin and ethidium bromide) interfere with the metabolic machinery of cells directly. Triethyltin, musk Tetralin@, hexa- chlorophene, isoniazid, and cyanide inhibit mitochondrial respiration (Cammer, 19801. Edematous swelling of the myelin sheath (often re- versible without demyelination and remyelination) is the hallmark of these toxic myelinopathies. Widespread spongiform degeneration of white matter is typical, and signs of increased intracranial pressure (e.g., headache, vertigo, visual disturbances, behavioral changes, con- vulsions, and coma) and increased CSF protein are common clinical features (Lane and Routledge, 1983; Powell et al., 19801. Changes in muscle induced by chemical agents (mostly drugs) are well known. Steroids commonly produce a proximal myopathy asso- ciated with weakness and wasting of the quadriceps muscle. Myopathies associated with muscle pain, tenderness, stiffness, and cramping in a proximal, largely symmetrical distribution are also relatively common (Lane and Routledge, 1983~. This syndrome may be associated with a necrotizing myopathy (e.g., from exposure to clofibrate), an inflam- matory myositis (e.g., in Spanish toxic oil syndrome), or a hypokalemic vacuolar myopathy (e.g., from exposure to barium salts). Rhab- domyolysis, an uncommon and occasionally fatal syndrome heralded by severe muscle pain and swelling, sometimes occurs in drug addicts, e.g., users of ethanol, opiates, or phencyclidine (PCP). Myatonia may develop from treatment with ,8-adrenergic receptor agonists. Other types of neurotoxicity are less well characterized. Changes in astrocytes have been induced experimentally by intoxication with water, ouabain, and antimetabolites (Powell et al., 19801. Capillary damage with extravasation of blood cells occurs in humans or animals with hemorrhagic encephalopathy induced by lead, cadmium, indium, or terbium (Jacobs, 1980~. Bismuth administered intramuscularly report- edly can precipitate spinal ischemia and spastic paraplegia (Sterman and Schaumburg, 1980~. A few agents (e.g., alkyl nitrosoureas) can induce primary malignant tumors in brain or nerves. Finally, important changes are believed to occur in the neuroendocrine system in response to selected toxicants such as DDT and chlordecone.

~ 24 DRINKING WATER AND H"LTH Susceptibilities of Special Populations THE AGED Aging is associated with significant changes in the nervous system (Katzman and Terry, 19831; age-related decline in neural function may allow previously silent neurotoxic disorders to reach clinical expression. Clinical experience with therapeutic drugs indicates that the elderly, es- pecially those with metabolic abnormalities or with hepatic or renal im- pairment, are more susceptible to the toxic effects of xenobiotic substances. However, systematic studies of the effects of chemical neurotoxicants on laboratory animals of varying age have yet to be undertaken. The most intensively studied and best documented aspects of normal human aging are changes in intellect and memory. Intellectual perfor- mance, as measured by tests of vocabulary, information retention, and comprehension, reaches a peak between ages 20 and 30 and is maintained throughout adult life, at least until the mid-70s, in the absence of disease. Perceptual processing and choice reaction time is slowed during aging. Learning, storage, and retrieval of information associated with short-term memory are consistently impaired in older subjects. Defective thermal regulation and decreased lacrimation may also occur. Sleeping patterns are altered. Motor tasks, including locomotion, handwriting, and other purposeful movements, are performed more slowly, weakly, or in an uncoordinated manner. Vibration sense is progressively impaired with advancing age, touch sensation is diminished, thermal discrimination is impaired, and pain threshold is mildly raised. Muscle wasting is common, strength is reduced, and tendon reflexes are difficult to elicit (Katzman and Terry, 1983~. All these changes may also occur as a consequence of chemical intoxication (Silverstein, 19821. Neurobehavioral changes accompanying human aging are associated with structural or functional alterations in the CNS and PNS. Some changes, such as alterations in vascular and cardiac reflexes, galvanic skin response, potency, micturition, and papillary response, probably result from changes in the autonomic nervous system (Katzman and Terry, 19831. Sympathetic hyperactivity is commonly present. This may interfere with cognitive functioning in older subjects, especially under the stress of psychological testing. Such individuals would be expected to be more susceptible to chemical toxicants with sympathomimetic properties. Changes in cerebral blood flow, essential for the maintenance of normal brain function, may occur over the age of 80 years. Cortical atrophy and ventricular enlargement have been documented in normal individuals during senescence. Certain regions of the brain are more susceptible to neuronal cell loss: the locus ceruleus and substantia

Neurotoxic Effects 125 nigra, both of which undergo maximum reduction in the third and fourth decades and slowly decline thereafter, Purkinje cells, and putamen neu- rons, which decline in number at a linear rate (Katzman and Terry, 19831. On the other hand, several cranial nuclei and the olivary nucleus maintain stable populations. The number of cerebral cortical neurons may be re- duced by one-half from age 20 to 80 years, and supplementary reductions and alterations occur in their dendritic arborizations and synaptic inputs. The volume of a yellow, insoluble pigment, lipofuscin, increases at a linear rate in most neurons with increasing age, but there is no evidence that this material is cytotoxic. Other types of neuronal pathology occur in normal aged brains, including neurofibrillary tangles, neuritic plaques, and granulovacuolar bodies. Neuronal changes and cell loss result in substantial local or more generalized alterations in the normal concentra- tion of neurotransmitters, including dopamine, nonepinephrine, serotonin, GABA, and choline acetyltransferase the enzyme required for the syn- thesis of the neurotransmitter acetylcholine (Katzman and Terry, 19831. Morphological changes in the PNS include a probable reduction of sensory neurons, an increase in the normal incidence of demyelination in spinal roots and peripheral nerves, increased amounts of connective tissue, and a mild loss of myelinated fibers. The central processes of dorsal root ganglion cells typically undergo distal dystrophic and degenerative changes (Spencer and Ochoa, 19811. There may be a slight reduction in the number of motor neurons with age, and regressive changes have been reported in the terminals of motor axons. Changes in sensory and motor nerve con- duction along with a progressive slowing of the nerve action potentials are also characteristic. THE DISEASED There are compelling theoretical reasons to suspect that people with certain diseases and under specific therapeutic regimens may be more susceptible than others to environmental neurotoxicants. This immense subject can only be considered briefly in this report. Many therapeutic drugs induce neurological disorders, and some of these might be potentiated (or suppressed) by concurrent exposure to environmental toxicants acting at related sites in the nervous system. Exposure to chemical substances may also unmask a latent neurological or neuromuscular disorder in previously asymptomatic individuals. Some unfavorable responses to therapy increase in frequency and severity in direct proportion to dose and duration of treatment; others appear to be unassociated with dose and probably result from hypersensitivity reactions (Lane and Routledge, 1983; Silverstein, 19821. Some drugs have the potential to act on the cerebral cortex to produce coma, seizures, or strokes.

]26 DRINKING WATER AND HEALTH Headache occurs as a result of exposure to agents that stretch pain-sensitive blood vessels and meninges. The basal ganglia are a target of many psychoactive or neuroactive drugs that cause a range of disabling effects, including tremor, asterixis, myoclonus, dyskinesia, Catalonia, and dys- tonia. Other drugs interfere with cranial nerve function, leading to an- osmia, trigeminal neuropathy, extraocular movement disorders and ataxia, oculotoxicity, or ototoxicity. When ototoxicity is associated with thera- peutic doses of aminoglycoside antibiotics, the extent of damage can be exacerbated by concurrent exposure to noise (Prosen and Stebbins, 19801. Finally, many widely used drugs produce neuromuscular disorders that are expressed as peripheral neuropathy, myasthenic syndromes, and var- ious types of myopathy. Neurological diseases of childhood and adulthood are numerous. Some of them are treated with therapeutic agents that also may have neurological or psychological side effects. Nutritional disorders and inadequate pre- ventive medicine may compound the situation in developing countries. Among the most common conditions associated with neurological dys- function in childhood are convulsive disorders. In adults, changes in neural structure or function commonly develop in association with such common syndromes as alcoholism, diabetes mellitus, and hypertension. Psychiatric disorders, such as schizophrenia, are primary dysfunctions of the brain, possibly associated with irregularities of CNS neurotransmission. Such individuals may be especially susceptible to chemical agents that further perturb affected transmitter systems. The aging human is most likely to develop devastating degenerative changes in the nervous system. These include, for example, senile de- mentia of the Alzheimer type, a syndrome of progressive mental deteri- oration, loss of memory, and impaired cognitive function associated with primary degeneration of neurons predominantly in the cerebral cortex and hippocampus; parkinsonism and disorders of motor function associated with striatal degeneration of the substantia nigra and characterized by flexed posture, rigidity, tremor, and mental and autonomic deficits; and amyotrophic lateral sclerosis, a disorder predominantly affecting motor neurons in the motor cortex and spinal cord. Axonal syndromes, such as mild polyneuropathy, are also common among the elderly. Diseases of myelin increase in frequency after the third decade: multiple sclerosis, most likely to affect women, is a progressive but remitting demyelinating disorder associated with impaired vision, nystagmus, tremor, ataxia, para- plegia, bladder dysfunction, and altered emotional responses (Katzman and Terry, 1983~. Special sense organs also undergo important degener- ative changes during the aging process. For example, hearing impairment is a major affliction of the elderly, affecting more than one-fourth of those past 65 years of age (Katzman and Terry, 19831. Most cases are idiopathic,

Neurotoxic Effects 127 often of genetic origin, and result from loss of sensory hair cells in the inner ear and involvement of the auditory nerve. Significant visual im- pairment occurs in 6% of individuals over age 65 and in 46% of those over 85. Senile macular degeneration, associated with changes in the retinal epithelium, accounts for a proportion of these cases; nonneuronal change (cataract, glaucoma) accounts for the balance. Age-associated de- clines also occur in taste sensation and olfaction. Selected toxic agents are able to mimic many of the clinical and patho- logical features described in some of the disorders listed above. For ex- ample, the parkinsonian syndrome is observed in workers chronically exposed to manganese ore or carbon disulfide and in persons intoxicated for relatively short periods with MPTP (Langston, 1985~. Certain types of motor neuron disorders have been associated with exposure to lead, and polyneuropathy follows overexposure to a number of occupational chemicals. Psychoses can be exacerbated by acute exposure to diisopro- pylfluorophosphate. In summary, therefore, the nervous system appears to have a number of systems that are vulnerable both to the aging process and to toxic chemicals, and a limited repertoire of neurobehavioral responses. There- fore, it is likely that aging populations with compromised neural structure and function, as well as reduced capacity for liver metabolism and renal clearance, are more susceptible to certain neurotoxic substances than are their younger adult counterparts. Other groups that may be especially susceptible to neurotoxic agents include those occupationally exposed to chemicals with neurotoxic prop- erties, persons with renal dysfunction, and those with skin conditions that increase dermal absorption of chemical agents. STUDIES IN HUMANS Occurrence of Neurological Disease Experience has demonstrated that direct and indirect chemical interfer- ence with the nervous system can perturb virtually any part of the neuraxis, and the resulting disorders can usually be classified in terms of the ana- tomical site affected and the clinical presentation. In general, the signs and symptoms of drug-induced neurological disorders are virtually indis- tinguishable from those seen in naturally occurring disease but are usually reversible if diagnosed early enough (Lane and Routledge, 1983; see Table 4-24. Neurotoxicity commonly accompanies prolonged therapeutic treat- ment with anticonvulsants, anticholinergics, neuroleptics, antiparkinson drugs, and antineoplastic drugs (Katcher et al., 19831. One of the most recently recognized iatrogenic neurotoxicities is the sensory neuropathy

|28 DRINKING WATER AND H"LTH TABLE 4-2 Some Common Clinical Manifestations of Human Neurotoxicitya Function Affected Manifestation Chemical Cognitive Intelligence loss Lead saltsb Learning decrements Lead salts b Memory dysfunction Anticholinesterases Sensory Irritability Carbon disulfide Apathy/lethargy Carbon monoxide A t t e n t i o n d i f fi c u l t y A n t i c h o l i n e s t e r a s e s Illusions, delusions Ergot Dementia Aluminum Depression, euphoria Ozocerite Stupor, coma Dicyclopentadiene Sensory, special Abnormalities of: Smell Cadmium Vision Organomercury Taste Selenium Audition Toluene Balance Methyl nitrite Somatosensory Skin senses (e.g., Trichloroethylene numbness, pain) Proprioception Acrylamide Motor Muscle weakness, paralysis Organophosphates Spasticity p-N-Oxalylamino-~-alanine Rigidity MPTP Tremor Chlordecone Dystonia Manganese Incoordination Organomercury Hyperactivity Lead salts Myoclonus Toluene Fasciculation Anticholinesterases Cramps Styrene Seizures, convulsions Acetonitrile Autonomic Abnormalities of: Sweating Acrylamide Temperature control Chlordane Gastrointestinal function Lead salts Appetite/body weight Dinitrobenzene Cardiovascular control 1-Nitrophenyl-3-(3- pyridylmethyl) urea Urination Dimethylaminopropionitrile Sexual function ,B-Chloroprene Immune system Myositis, vasculitis, fibrosis Guillain Barre syndrome Spanish toxic oil Gold salts aAdapted from Lane and Routledge, 1983. bData not conclusive.

Neurotoxic Effects 129 syndrome associated with pyridoxine megavitamin therapy (Foca, 1985), prescribed for the treatment of premenstrual tension. In developing countries, biological toxins (e.g., Clostridium botulinum, C. tetani, and Corynebacterium diphtherial, neurotoxic agents naturally present in food (e.g., cassava and Lathyrus suppl.) or as contaminants (e.g., ergot and aflatoxin), and pesticides probably account for a large number of human neurotoxic disorders. Uncontrolled cholinergic crises, sometimes leading to death, are commonplace in certain regions among agricultural and pesticide workers (Almeida, 1984), and long-lasting changes in the electroencephalograms and behavior of surviving persons have been recorded (Duffy et al., 19791. Other pesticides contain tremor- and seizure- inducing organochlorines or synthetic pyrethroids that perturb neurotrans- mission (Narahashi, 1984; Taylor et al., 19791. The worldwide problem of substance abuse, particularly abuses involving ethanol, hallucinogens, narcotics, CNS stimulants, solvents, and nitrous oxide, leads to various types of short- or long-lasting neurological dysfunction. Many other sub- stances encountered in the workplace (e.g., solvents, monomers, and catalysts) have been associated with neurological illnesses ranging from polyneuropathy to organic brain syndrome. Both of these disorders were observed in workers exposed for only a few weeks to one particularly potent industrial neurotoxicant, Lucel-7 (2-tert-butylazo-2-hydroxy-5- methylhexane) (Kurt and Webb, 19801. Among the environmental pollutants with neurotoxic potential, lead and mercury each occupy a prominent position, although the number of people in North America with overt neurotoxic disorders attributable to these sources is probably low. Potent chemical toxins are secreted by or con- tained in numerous members of the animal kingdom. Many of these agents disturb nerve conduction, and one, ciguatoxin, is a major cause of acute neurotoxicity in the Pacific among those who eat contaminated fish (Kap- lan, 19801. Others (e.g., c~-bungarotoxin and cx-latrotoxin) affect synaptic transmission. Both types can produce acute, life-threatening conditions. Some of these agents find their way into food and water consumed by humans. Some chemicals with experimentally proven neurotoxic potential in animals are used as food additives (e.g., monosodium glutamate), flavors and fragrances (e.g., 2,6-dinitro-3-methoxy-4-tert-butyltoluene), and antiseborrheic agents (e.g., zinc pyridinethione), but no cases of human neurotoxic disease from these sources have been reported. The occurrence of subclinical neurological and behavioral disorders associated with chemical substances is unknown but is believed by some to be widespread. Examples include the unre~.olved controversies about childhood cognitive impairment from environmental lead contamination and the neurobehavioral effects attributed to prolonged occupational ex- posure to a variety of industrial solvents.

]30 DRINKING WATER AND HEALTH Methods for assessing neurotoxic diseases in humans fall outside of the scope of this document. For information on this subject, please refer to Geller et al. (1979) and Spencer and Schaumburg (1980, pp. 650-7071. Epidemiological Studies There have been several epidemiological studies of outbreaks of human neurobehavioral disorders in which chemical compounds have been im- plicated and subsequently proved to be neurotoxic in animal studies. Few of these investigations have been focused on features important to the toxicologist, such as the exposure concentration and duration, route of entry, and individual susceptibility. A notable exception was the outbreak of neuropathy induced by methyl n-butyl ketone (MBK) in an Ohio factory in 1973 (Allen et al., 1975), where an analysis of the atmosphere and exposure conditions was conducted by NIOSH. This investigation was followed by a large number of experimental animal studies that demon- strated, for example, that dermal penetration was a significant route of human exposure to MBK, that oxidative metabolism yielded a number of metabolites with neurotoxic potency, and that other solvents played a role in potentiating the neurotoxic activity of MBK (Spencer and Schaumburg, 1980~. Other notable recent outbreaks subjected to epidemiological investi- gation include occupational injuries associated with exposures to chlor- decone, dimethylaminopropionitrile, n-hexane, Lucel-7, and leptophos (Pestronk et al., 1979, 1980; Spencer et al., 1980a,b, 1985; Taylor et al., 1978; Xintaras et al., 1978~. Other major outbreaks have involved con- sumption of foodstuffs contaminated with such neurotoxic substances as hexachlorobenzene, tri-o-cresyl phosphate, methylmercury, polybromi- nated biphenyls, or thallium salts (Schaumburg and Spencer, 19801. A small outbreak of neuropathy and encephalopathy in a Japanese family followed consumption of well water contaminated with acrylamide (Igisu et al., 19751. CONTROLLED AND OTHER STUDIES Controlled studies of the human response to environmental neurotoxi- cants have been limited to acute exposure associated with short-term and presumably reversible neurobehavioral dysfunction. Some of these studies have been focused on the effects of inhaled organic solvents on manual dexterity, performance on psychological test batteries, vestibular func- tions, and other functions. Another set of data from relatively controlled situations reflects clinical experience with therapeutic drugs. Although many of these compounds are not of immediate concern, some agents

Neurotoxic Effects 131 (e.g., anticholinesterases) used in the therapeutic setting may also appear in different forms in drinking water. STUDIES IN ANIMALS Many methods have been used to assess neurotoxicity in animals, and many types of data have been produced. Experimental studies of the adverse actions of chemical agents on the nervous system include system- atic observation and measurement of behavior, neural function, structure, and biochemistry in various laboratory animals. Behavioral toxicology and teratology are relatively recent introductions to the field, and the specific methods used in these studies are evolving rapidly. In this section, no attempt has been made to describe methodology in detail. In broad terms, behavioral toxicology deals with two groups: respondent behaviors (e.g., auditory startle response), which are elicited by a specific observable stimulus and determined primarily by the properties of the stimulus presentation, and operant behaviors (e.g., exploratory activity), which occur in the absence of an eliciting stimulus and are determined primarily by their consequences. Operant behavior is frequently condi- tioned by reward or punishment to generate a reliable quantitative baseline. Some of the test methods used in behavioral toxicology, such as the psychophysical measurement of cortical blindness induced by methyl- mercury, have been successfully used to model visual neurotoxicity in humans, whereas the findings of other tests bear an unknown relationship to human disease. However, with further refinement of methods and the ability to evaluate specific changes in relation to biochemical or functional changes in the nervous system, behavioral toxicology and teratology should be especially useful in studying human neurobehavioral disorders (such as tardive dyskinesia), which appear to be unassociated with detectable neuropathological changes. A variety of electrophysiological techniques have been used to assess the presence and extent of CNS and PNS injuries, to locate the target sites of action, and to determine the cellular and molecular mechanisms of action. These techniques include measurements of PNS sensory and motor conduction velocities, tests of sensory- and motor-evoked cortical poten- tials, electroencephalography, electromyography, and extracellular unit recordings from retinal ganglion cells. Intracellular recordings, voltage and patch clamping, and noise analysis have also been used (Fox et al., 1982; Narahashi, 19841. Several of these experimental techniques have been used in clinical situations, since they are noninvasive, sensitive, reliable, and relatively easy to use. These experimental and clinical tech- niques thus allow one to determine the contribution of pathophysiology to neurological and behavioral impairment.

]32 DRINKING WATER AND HEALTH In experimental animals, observation and measurement of changes in neural structure have provided a solid basis for understanding human neurotoxic syndromes associated with pathological changes in the devel- oping or adult nervous system. This has been possible because the methods used to assess neurological disorders in humans have been exploited and further refined by experimental neurotoxicologists. The transmission elec- tron microscope has greatly extended the resolution of tissue detail, which used to be afforded by the light microscope. Armed with these techniques and a knowledge of neurobiological principles, one can define different patterns of neurotoxic response to chemical agents and explain how these result in human diseases that vary in expression and prognosis. With the exception of certain developmental disorders associated with chemical overexposure, there is usually an excellent correlation between the neu- rotoxic disorders of mammals and human beings. Through studies in various animal models, structural and functional changes induced by nu- merous chemical substances have been characterized. These studies have also enabled investigators to pinpoint the initial site of neurotoxic damage or dysfunction information that provides a sensitive method to assess no-effect and threshold levels for the substance of interest. With few exceptions, biochemical mechanisms underlying the neuro- toxic action of environmental chemicals are poorly understood. The best examples are the actions of certain organophosphate esters on acetylcho- linesterase, which promptly induce cholinergic toxicity, and on neuropathy target esterase (formerly called neurotoxic esterase), which is associated with the development of polyneuropathy (Davis and Richardson, 1980~. From measurements of enzyme levels in tissue and blood, one can estimate the degree of enzyme inhibition and associated neurotoxic illness. Such measurements provide a sensitive means of assessing the neurotoxic ac- tivities of agents with these properties. Biochemical studies of other dis- orders will most likely yield information on mechanisms that will, in turn, provide comparably sensitive biological markers to determine the degree of neurotoxic impairment produced by chemical agents. Data on bio- chemical changes underlying neurotoxicity also are critical for the pre- vention and treatment of such disorders. Information correlating metabolism and pharmacodynamics with neu- rotoxic properties is also sparse. The usefulness of this approach, however, was elegantly demonstrated in studies of n-hexane and methyl n-butyl ketone in rats. O'Donoghue and colleagues (1982) found that these chem- ically distinct solvents appeared to produce qualitatively identical patterns of neurotoxic disease. The neurotoxic potency of each of the agents was directly proportional to the amount of y-diketone generated~by metabolism. Such studies in rats and subsequently in humans (DiVincenzo et al., 1978; Perbellini et al., 1979, 1980) demonstrated that these agents were meta-

Neurotoxic Effects 133 helically related and each could be converted to the proximal metabolite 2,5-hexanedione. Moreover, the onset of characteristic neurological dys- function (e.g., hindlimb weakness) was linearly related to the concentra- tion of the proximal metabolite in serum (Krasavage et al., 19801. How many other neurotoxic compounds show a linear relationship between dose and effect is unknown. I NTERSPECI ES EXTRAPOLATION There are often major differences between the degree of neurotoxic response observed in animals and that found in humans. For example, the rat is relatively refractory to such agents as organophosphates, arsenic, thallium, MPTP, and thalidomide, all of which readily produce neuro- degenerative disorders in humans. Other substances that cause neuropathy in humans, such as acrylamide (see Chapter 9), n-hexane (Spencer et al., 1980a), and carbon disulfide (Seppalainen and Haltia, 1980), also produce neuropathy in rodents. Understanding in this area is so limited that there is no reliable prescription for choosing the appropriate animal for studies intended to shed light on human susceptibility. One important exception is the choice of fowl for the study of organophosphate neuropathy (Davis and Richardson, 1980~. Once a suitable test animal has been identified and a neurotoxic dose established, then the disorder of interest will usually appear in all adult animals after a similar dose and test period have elapsed. The pattern of dysfunction and the underlying structural and functional changes are also uniform across test animals within a single species. These principles may allow the experimentalist to test much smaller numbers of animals than those used, for example, in assays of carcinogenesis. Caution is appro- priate, however, in studies focused on the establishment of threshold or low-risk levels of neurotoxic response (as is generally the case in risk assessment), where one is operating in a portion of the dose-response curve that is likely to be less steep than the region in which most exper- imental studies have been performed. These rules do not apply to neu- rotoxic disorders of the developing animal, in which response rate and type of teratogenicity are often variable. RISK ASSESSMENT Little effort has been directed toward assessing risk from exposure to chemical neurotoxicants. Because of the large number of potential health end points associated with neurotoxicants, attempts at risk assessment are substantially more complex than they are for chemical carcinogens. For example, the same substance may produce markedly different neurotoxic

]34 DRINKING WATER AND HEALTH effects, depending on age at time of exposure. Although the susceptibility of the nervous system to toxicity is well recognized, no bioassay has been developed for identifying and regulating environmental neurotoxicants. Assessment of dose-response relationships is plagued by the usual un- certainties. Epidemiological dose-response information is extremely lim- ited, and that which does exist may not be applicable to the general population. Although some animal models are particularly accurate for assessing neurological disorders in humans, the large number of potential toxic end points greatly complicates assessment since there may not be a single threshold dose for all end points. Exposure assessment requires knowledge concerning chemical type and concentration, effect and its seventy, route and duration of exposure, and identification of high-nsk groups. Safety factors used in such assessments should reflect uncertainty due to variability within the population and variability between the population of interest and the sampled population. Until an accurate probabilistic model for risk assessment is developed, safety factors will have to be used in place of confidence limits. A more detailed discussion of risk assessment for neurotoxicants is found in Chap- ter 8. REFERENCES Allen, N., J. R. Mendell, D. J. Billmaier, R. E. Fontaine, andJ. O'Neill. 1975. Toxic polyneuropathy due to methyl n-butyl ketone. An industrial outbreak. Arch. Neurol. 32:209-218. Almeida, W. F. 1984. The dangers and the precautions. World Health Aug./Sept.:10-12. Altenkirch, H., H. M. Wagner, G. Stoltenburg, and P. S. Spencer. 1982. Nervous system responses of rats to subchronic inhalation of n-hexane and n-hexane + methyl-ethyl- ketone mixtures. J. Neurol. Sci. 57:209-219. Anger, W. K. 1984. Neurobehavioral testing of chemicals: Impact on recommended stan- dards. Neurobehav. Toxicol. Teratol. 6:147-153. Anger, W. K., and B. L. Johnson. 1985. Chemicals affecting behavior. Pp. 51-148 in J. L. O'Donoghue, ed. Neurotoxicity of Industrial and Commercial Chemicals. Vol. 1. CRC Press, Boca Raton, Fla. Brady, S. T. 1984. Basic properties of fast axonal transport and the role of fast transport in axonal growth. Pp. 13-29 in J. S. Elam and P. Cancalon, eds. Axonal Transport in Neuronal Growth and Regeneration. Plenum, New York. Cammer, W. 1980. Toxic demyelination: Biochemical studies and hypothetical mecha- nisms. Pp. 239-256 in P. S. Spencer and H. H. Schaumburg, eds. Experimental and Clinical Neurotoxicology. Williams & Wilkins, Baltimore. Cho, E. S., P. S. Spencer, B. S. Jortner, and H. H. Schaumburg. 1980. A single intravenous injection of doxorubicin (Adriamycin~) induces sensory neuropathy in rats. Neurotox- icology 1 :583-591. Cooper, J. R., F. E. Bloom, and R. H. Roth. 1982. The Biochemical Basis of Neuro- pharmacology, 4th ed. Oxford University Press, New York. 367 pp.

Neurotoxic Effects 135 Damstra, T., and S. C. Bondy. 1980. The current status and future of biochemical assays for neurotoxicity. Pp. 820-833 in P. S. Spencer and H. H. Schaumburg, eds. Experimental and Clinical Neurotoxicology. Williams & WiLkins, Baltimore. Davis, C. S., and R. J. Richardson. 1980. Organophosphorus compounds. Pp. 527-544 in P. S. Spencer and H. H. Schaumburg, eds. Experimental and Clinical Neurotoxicology. Williams & WiLkins, Baltimore. DiVincenzo, G. D., M. L. Hamilton, C. J. Kaplan, W. J. Krasavage, and J. L. O'Don- oghue. 1978. Studies on the respiratory uptake and excretion and the skin absorption of methyl n-butyl ketone in humans and dogs. Toxicol. Appl. Pharmacol. 44:593-604. Duffy, F. H., J. L. Burchfiel, P. H. Bartels, M. Gaon, and V. M. Sim. 1979. Long-term effects of an organophosphate upon the human electroencephalogram. Toxicol. Appl. Pharmacol. 47: 161-176. Ells, J. 1984. Drugs used in tuberculosis and leprosy. Pp. 287-291 in M. N. G. Dukes and J. Ells, eds. Side Effects of Drugs Annual 8. A Worldwide Yearly Survey of New Data and Trends. Elsevier, New York. Foca, F. J. 1985. Motor and sensory neuropathy secondary to excessive pyridoxine inges- tion. Arch. Phys. Med. Rehab. 66:634-636. Fox, D. A., H. E. Lowndes, and G. G. Bierkamper. 1982. Electrophysiological techniques in neurotoxicology. Pp. 299-335 in C. L. Mitchell, ed. Nervous System Toxicology. Raven Press, New York. Geller, I., W. C. Stebbins, and M. J. Wayner, eds. 1979. Test Methods for Definition of Effects of Toxic Substances on Behavior and Neuromotor Function. Proceedings of the workshop held April 1-4, 1979, San Antonio, Texas. Report No. EPA 560/11-79-010. (Neurobehavioral Toxicology, Vol. 1, Suppl. 1.) ANKHO International Inc., Fayette- ville, N.Y. 225 pp. Ginsberg, M. D. 1980. Carbon monoxide. Pp. 374-394 in P. S. Spencer and H. H. Schaumburg, eds. Experimental and Clinical Neurotoxicology. Williams & Wilkins, Baltimore. Gopalan, C. 1950. The lathyrism syndrome. Trans. R. Soc. Trop. Med. Hyg. 44:333-338. Igisu, H., I. Goto, Y. Kawamura, M. Kato, K. Izumi, and Y. Kuroiwa. 1975. Acrylamide encephaloneuropathy due to well water pollution. J. Neurol. Neurosurg. Psych. 38:581- 584. Jacobs, J. M. 1980. Vascular permeability and neural injury. Pp. 102-117 in P. S. Spencer and H. H. Schaumburg, eds. Experimental and Clinical Neurotoxicology. Williams & Wilkins, Baltimore. Jacobson, M. 1978. Developmental Neurobiology, 2nd ed. Plenum, New York. 562 pp. Johnson, M. K. 1975a. Structure-activity relationships for substrates and inhibitors of hen brain neurotoxic esterase. Biochem. Pharmacol. 24:797-805. Johnson, M. K. 1975b. The delayed neuropathy caused by some organophosphorus esters: Mechanism and challenge. CRC Crit. Rev. Toxicol. 3:289-316. Kandel, E. R., and J. H. Schwartz, eds. 1981. Principles of Neural Science. Elsevier, New York. 733 pp. Kaplan, J. G. 1980. Neurotoxicity of selected biological toxins. Pp. 631-648 in P. S. Spencer and H. H. Schaumburg, eds. Experimental and Clinical Neurotoxicology. Wil- liams & WiLkins, Baltimore. Katcher, B. S., L. Y. Young, and M. A. Koda-Kimble, eds. 1983. Applied Therapeutics: The Clinical Use of Drugs, 3rd ed. Applied Therapeutics, San Francisco. 1619 pp. Katzman, R., and R. Terry, eds. 1983. The Neurology of Aging. F. A. Davis, Philadelphia. 249 pp.

t36 DRINKING WATER AND HEALTH Krasavage, W. J., J. L. O'Donoghue, G. D. DiVincenzo, and C. J. Terhaar. 1980. The relative neurotoxicity of methyl-n-butyl ketone, n-hexane and their metabolites. Toxicol. Appl. Pharmacol. 52:433-441. Kristensson, K. 1984. Retrograde signaling after nerve injury. Pp. 31-43 in J. S. Elam and P. Cancalon, eds. Axonal Transport in Neuronal Growth and Regeneration. Plenum, New York. Kurt, T., and C. R. Webb, Jr. 1980. Toxic occupational neuropathy—Texas. Morbid. Mortal. Weekly Rep. 29:529-530. Lane, R. J. M., and P. A. Routledge. 1983. Drug-induced neurological disorders. Drugs 26: 124-147. Langston, J. W. 1985. MPTP neurotoxicity: An overview and characterization of phases of toxicity. Life Sci. 36:201-206. Le Quesne, P. M. 1980. Acrylamide. Pp. 309-325 in P. S. Spencer and H. H. Schaumburg, eds. Experimental and Clinical Neurotoxicology. Williams & Wilkins, Baltimore. Mayer, S. E. 1980. Neurohumoral transmission and the autonomic nervous system. Pp. 56-90 in A. G. Gilman, L. S. Goodman, and A. Gilman, eds. Goodman and Gilman's The Pharmacological Basis of Therapeutics, 6th ed. Macmillan, New York. McLeod, J. G. 1983. Distal autonomic neuropathy. Pp. 543-573 in R. Bannister, ed. Autonomic Failure: A Textbook of Clinical Disorders of the Autonomic Nervous System. Oxford University Press, New York. Merigan, W. H., and B. Weiss. 1980. Neurotoxicity of the Visual System. Raven Press, New York. 274 pp. Narahashi, T. 1984. Nerve membrane sodium channels as the targets of pyrethroids. Pp. 85-108 in T. Narahashi, ed. Cellular and Molecular Neurotoxicology. Raven Press, New York. NIOSH (National Institute for Occupational Safety and Health). 1977. National Occupa- tional Hazard Survey. Vol. III. Survey Analysis and Supplemental Tables. DHEW (NIOSH) Publication No. 78-114. National Institute for Occupational Safety and Health, Cincinnati. 799 pp. (Available from National Technical Information Service, Springfield, Va., as Publication No. PB-82-229881.) NRC (National Research Council). 1982. Possible Long-Term Health Effects of Short- Term Exposure to Chemical Agents. Vol. 1. Anticholinesterases and Anticholinergics. National Academy Press, Washington, D.C. 292 pp. O'Donoghue, J. L. 1985. Neurotoxicity of Industrial and Commercial Chemicals. Vol. II. CRC Press, Boca Raton, Fla. 209 pp. O'Donoghue, J. L., W. J. Krasavage, G. D. DiVincenzo, and D. A. Ziegler. 1982. Commercial-grade methyl heptyl ketone (5-methyl-2-octanone) neurotoxicity: Contri- bution of 5-nonanone. Toxicol. Appl. Pharmacol. 62:307-316. Olney, J. W. 1980. Excitotoxic mechanisms of neurotoxicity. Pp. 272-294 in P. S. Spencer and H. H. Schaumburg, eds. Experimental and Clinical Neurotoxicology. Williams & WiLkins, Baltimore. OTA (Office of Technology Assessment). 1984. Impacts of Neuroscience A Background Paper. Publication No. OTA-BP-BA-24. Off~ce of Technology Assessment, Congress of the United States, Washington, D.C. 40 pp. Pansky, B., and D. J. Allen. 1980. Pp. 10-11 in Review of Neuroscience. Macmillan, New York. Perbellini, L., F. Brugnone, G. Pastorello, and L. Grigolini. 1979. {Jrinary excretion of n-hexane metabolites in rats and humans. Int. Arch. Occup. Environ. Health 42:349- 354.

Neurotoxic Effects 137 Perbellini, L., F. Brugnone, and I. Pavan. 1980. Identification of the metabolites of n- hexane, cyclohexane, and their isomers in men's urine. Toxicol. Appl. Pharmacol. 53:220-229. Pestronk, A., J. P. Keogh, and J. W. Griffin. 1979. Dimethylaminopropionitrile (DMAPN) intoxication: A new industrial neuropathy. (Abstract GS-1.) Neurology 29:540. Pestronk, A., J. P. Keogh, and J. W. Griffin. 1980. Dimethylaminopropionitrile. Pp. 422- 429 in P. S. Spencer and H. H. Schaumburg, eds. Experimental and Clinical Neuro- toxicology. Williams & Wilkins, Baltimore. Powell, H. C., R. R. Myers, and P. W. Lampert. 1980. Edema in neurotoxic injury. Pp. 118-138 in P. S. Spencer and H. H. Schaumburg, eds. Experimental and Clinical Neu- rotoxicology. Williams & Wilkins, Baltimore. Price, D. L., J. Griffin, A. Young, K. Peck, and A. Stocks. 1975. Tetanus toxin: Direct evidence for retrograde intraaxonal transport. Science 188:945-947. Price Evans, D. A. 1978. Genetic studies involving drug metabolism in man. Pp. 135-155 in J. W. Gorrod and A. H. Beckett. Drug Metabolism in Man. Taylor & Francis, London. Prosen, C. A., and W. C. Stebbins. 1980. Ototoxicity. Pp. 62-76 in P. S. Spencer and H. H. Schaumburg, eds. Experimental and Clinical Neurotoxicology. Williams & Wil- kins, Baltimore. Rasminsky, M. 1980. Physiological consequences of demyelination. Pp. 257-271 in P. S. Spencer and H. H. Schaumburg, eds. Experimental and Clinical Neurotoxicology. Wil- liams & Wilkins, Baltimore. Rebert, C. S., S. S. Sorenson, R. A. Howd, and G. T. Pryor. 1982. Toluene-induced hearing loss in rats evidenced by the brainstem auditory-evoked response. Pp. 314-319 in H. N. MacFarland, C. E. Holdsworth, J. A. MacGregor, R. W. Call, and M. L. Kane, eds. The Toxicology of Petroleum Hydrocarbons. Proceedings of the Symposium, May 1982, Washington, D.C. American Petroleum Institute, Washington, D.C. Schaumburg, H. H., and P. S. Spencer. 1980. Selected outbreaks of neurotoxic disease. Pp. 883-889 in P. S. Spencer and H. H. Schaumburg, eds. Experimental and Clinical Neurotoxicology. Williams & Wilkins, Baltimore. Schaumburg, H. H., P. S. Spencer, and P. K. Thomas. 1983. Pp. 127-128 in Disorders of Peripheral Nerves. F. A. Davis, Philadelphia. Seppalainen, A. M., and M. Haltia. 1980. Carbon disulfide. Pp. 356-373 in P. S. Spencer and H. H. Schaumburg, eds. Experimental and Clinical Neurotoxicology. Williams & Wilkins, Baltimore. Shils, M. E., and E. V. McCollum. 1943. Further studies on the symptoms of manganese deficiency in the rat and mouse. J. Nutr. 26:1-19. Silverstein, A., ed. 1982. Neurological Complications of Therapy. Futura, New York. 470 PP Spencer, P. S., and J. Ochoa. 1981. The mammalian peripheral nervous system in old age. Pp. 35-103 in J. E. Johnson, Jr., ed. Aging and Cell Structure. Vol. I. Plenum, New York. Spencer, P. S., and H. H. Schaumburg, eds. 1980. Experimental and Clinical Neurotox- icology. Williams & Wilkins, Baltimore. 929 pp. Spencer, P. S., and H. H. Schaumburg. 1984. An expanded classification of neurotoxic responses based on cellular targets of chemical agents. Acta Neurol. Scand. 70(Suppl. 100):9-20. Spencer, P. S., D. Couri, and H. H. Schaumburg. 1980a. n-Hexane and methyl n-butyl ketone. Pp. 456-475 in P. S. Spencer and H. H. Schaumburg, eds. Experimental and Clinical Neurotoxicology. Williams & Wilkins, Baltimore.

|38 DRINKING WATER AND H"LTH Spencer, P. S., H. H. Schaumburg, M. I. Sabri, and B. Veronesi. 1980b. The enlarging view of hexacarbon neurotoxicity. CRC Crit. Rev. Toxicol. 7:279-356. Spencer, P. S., C. M. Beaubernard, M. C. Bischoff-Fenton, and T. L. Kurt. 1985. Clinical and experimental neurotoxicity of 2-t-butylazo-2-hydroxy-5-methylhexane. Ann. Neurol. 17:28-32. Sterman, A. B., and H. H. Schaumburg. 1980. The neurological examination. Pp. 675- 680 in P. S. Spencer and H. H. Schaumburg, eds. Experimental and Clinical Neuro- toxicology. Williams & Wilkins, Baltimore. Taylor, J. R., J. B. Selhorst, S. A. Houff, and A. J. Martinez. 1978. Chlordecone intox- ication in man. I. Clinical observations. Neurology 28:626-630. Taylor, J. R., V. P. Calabrese, and R. V. Blanke. 1979. Organochlorine and other insec- ticides. Pp. 391-455 in P. J. Vinken and G. W. Bruyn, eds. Handbook of Clinical Neurology. Vol. 36. Intoxications of the Nervous System. Part I. North-Holland, Am- sterdam. Thomas, P. K., D. J. Bradley, W. A. Bradley, P. H. Degen, G. Krinke, J. Muddle, H. H. Schaumburg, P. N. Skelton-Stroud, P. Thomann, and E. Tzebelikos. 1984. Correlated nerve conduction, somatosensory evoked potential and neuropathological studies in clioquinol and 2,5-hexanedione neurotoxicity in the baboon. J. Neurol. Sci. 64:277-295. Thomas, P. K., and J. Ochoa. 1984. Microscopic anatomy of peripheral nerve fibers. Pp. 39-96 in P. J. Dyck, P. K. Thomas, E. H. Lambert, and R. Bunge, eds. Peripheral Neuropathy. Vol. I. W. B. Saunders, Philadelphia. Varon, S. S., and G. G. Somjen. 1979. Glial typology and gliogenesis. Neurosci. Res. Prog. Bull. 17:19-41. Vorhees, C. V., and R. E. Butcher. 1982. Behavioural teratogenicity. Pp. 247-298 in K. Snell, ed. Developmental Toxicology. Praeger, New York. Weller, R. O., M. Swash, D. L. McLellan, and C. L. Scholtz. 1983. Clinical Neuro- pathology. Springer-Verlag, Berlin. 329 pp. Wilson, J. G., and F. C. Fraser. 1977a. Handbook of Teratology. Vol. 1. General Principles and Etiology. Plenum, New York. 476 pp. Wilson, J. G., and F. C. Fraser. 1977b. Handbook of Teratology. Vol. 2. Mechanisms and Pathogenesis. Plenum, New York. 491 pp. Wilson, J. G., and F. C. Fraser. 1978. Handbook of Teratology. Vol. 4. Research Pro- cedures and Data Analysis. Plenum, New York. 458 pp. Xintaras, C., J. R. Burg, S. Tanaka, S. T. Lee, B. L. Johnson, C. A. Cottrill, and J. Bender. 1978. NIOSH Health Survey of Velsicol Pesticide Workers. Occupational Ex- posure to Leptophos and Other Chemicals. DHEW (NIOSH) Publication No. 78-136. National Institute for Occupational Safety and Health, U.S. Department of Health, Ed- ucation, and Welfare, Cincinnati. 148 pp.