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.
26 Morphologic, Neurochemical, and Behavioral Responses to Toxic Agents This chapter discusses the importance of timing and dose for effects in develop- ing organisms. Then, specific biologic markers of neurodevelopment are dis- cussed, ranging from the association of minor physical anomalies with behavioral effects, to measurement of neurochemical concentrations, to behavioral assessment of complex processes. EFFECTS OF TIME OF EXPOSURE AND DOSE: IRRADIATION AS A PARADIGM The extensive use of ionizing radiation as an embryotoxic agent has been of para- mount importance in the delineation of several important concepts that have par- ticular relevance to current efforts in developmental toxicology. Studies in radiation teratology are unique, in that the physics of ionizing radiation has al- lowed scientists to produce effects in the embryo directly, with no concern for the moderating influence of the so-called placental barrier. Work from four laboratories established two basic concepts: stage specificity and the relation of dosage to response (Hicks, 1953; Rugh, 1953; Russell and Russell, 1954; Wilson, 1954~. All the early studies provided convincing evidence that the production of specific congenital malfor- 281 mations depended on the stage of develop- ment at which radiation was administered and that the severity of the effect pro- duced was a function of the dosage. The Reimplantation stages of development were usually shown to be relatively radio- resistant when neonatal death was as- sessed, although Rugh (1959) reported that low doses of radiation during this period produced a low incidence of exencephaly in survivors. Wilson (1954) showed the influence of dose, especially on the production of eye defects in rat fetuses. Irradiation with 25 reds on ED 9 produced eye defects in 6% of the fetuses. The magnitude of response increased with increasing dose; exposure to 200 reds produced eye defects in 72% of the fetuses. The embryo, under the experi- mental conditions, became radioresistant with time. Irradiation on ED 11 with either 25, 50, or 100 reds produced no eye defects in survivors, although all survivors were affected when the dose was 200 reds. Hicks ( 1953) also showed time-dependent re- sponses to ionizing radiation in studies of specific CNS defects. Hicks noted that irradiation of the embryo early in develop- ment produced forebrain defects, whereas irradiation during fetal and neonatal life resulted primarily in cerebellar malformations. As summarized by Rugh (1953), "the inert
282 , p'rimordium or the totally differentiated cell will be relatively resistant in terms of morphologic change. The actively dif- ferentiating intermediate stage or stages will be highly radiosensitive since they are in the process of transfor- mation Differentiation. Hicks devel- oped a mechanistic framework to validate the concept (Hicks, 1959; Hicks and D,Ama- to, 1966~. He hypothesized that the radio- sensitive cells (neuroblasts3 were local- ized to the region of the neural epithelium that was active in DNA synthesis and that the immediate consequence of ionizing radiation was massive and extensive death of cells of that type. Cell death was a transient phenomenon, and the embryo was capable of repairing the damage quickly. He postulated that the malformations ob- served at term were the consequence of a balance between the initial damage and the degree of regulation and regeneration inherent in that region of the embryo at the time of injury. Kallen (1965) documented that the devel- oping nervous system is capable of stage- dependent regeneration. The idea could be validated under conditions in which regenerative capacity was absent and radi- ation would produce specific cell death and specific neuronal deficits. As docu- mented below, the microneurons of the cere- bellum and the dentate gyrus, Which develop postnatally, satisfy the neces- sary conditions, and selective radiation of these regions at appropriate periods of development produces specific struc- tural and functional deficits. MICRONEURONAL RADIATION Cerebellum Hicks et al. (1962) reported that irradi- ation of the cerebellum of the 6-day-old rat with 200 R produced extensive damage to the external granule cell layer. The layer retained its regenerative capacity and formed an ectopic granule cell layer within 4 days. However, the initial damage also interfered with Purkinje cells: their form was altered, and the association of Purkinje cells and granule cells formed later was abnormal. Altman and coworkers NEURODEVELOPMENTAL TOXICOLOGY (Altman et al., 1969; Pellegrino and Alt- man, 1979) confirmed that observation in a series of experiments that effectively verified the idea that cell killing without substantial repair led to severe cellular deficits within the cerebellum. They used the knowledge that different microneuron- al populations that arise in the external granule layer do so at specific times after birth; the basket cells were formed on PN 6-7, the stellate cells on PN 8-11, and the granule cells on PN 8-21. They used sequen- tial irradiation to monitor morphogenesis and later performance in a variety of tasks. In the first group, focal irradia- tion of the cerebellar cortex with 200 R on PN 4 and 5 produced cerebellar disor- ganization similar to that observed by Hicks et al. (1962~. In a second group of animals, the focal irradiation consisted of 200 R on PN 4-5 (as above) followed by 150 R on PN 7, 9, 11, 13, and 15. With the fractionated regimen, all derivatives of the external granule cell layer failed to form-an effect that resulted in severe motor deficits. Postponing the initial irradiation to PN 8 and 12 produced selec- tive neu renal deficits (stellate and late - forming granule cells in one group, late- forming granule cells in another group) with corresponding selective behavioral effects. Initial observations failed to show any postural or motor deficits in the sec- ond group of rats that were irradiated many times postnatally (Pellegrino and Altman, 1979~. In fact, in experiments with a motor-driven rotating rod, the ir- radiated animals performed better than controls; that is, they fell off the rod less frequently. However, in open-field tests, the irradiated rats were observed to be significantly more active than con- trol animals. Ambulation in the open field is affected by agitation. It was concluded that microneuronal hypoplasia in the cerebellum that does not produce demonstrable locomotor deficits can nev- ertheless lead to hyperactivity at an age when the animals tend to be the most active (2 months). In adult animals, the differ- ence disappeared. The experiments documented a strong correlation between the developmental
RESPONSES TO TOXIC AGENTS history of a neuronal population and its contribution to the behavioral hier- archies within the animal. The studies are a logical extension of the classical studies reviewed earlier and confirm the observation that the primary effect of irradiation of sensitive cell populations is cell death. It is clear, however, that fractionated irradiation of developing microneurons is not accompanied by exten- sive regeneration. Hence, highly specific and highly reproducible cell deficits can be produced, and their behavioral conse- quences can be monitored. H. Ippocampus The same approach has been used to evalu- ate the effects of microneuronal hypoplas- ia on the hippocampus (Altman, 1986~. Focal x irradiation of the hippocampus, begun immediately after birth, prevents the formation of nearly 85% of the granule cells of the dentate gyrus. The rats were then tested in the same 283 postulated that these, and other, selec- tive effects on microneuronal populations can provide the anatomic basis for mini- mal brain dysfunction under the influence of a broad spectrum of environmental factors, such as alcohol, lead, and gluco- corticoids. APPLICATION TO OTHER TOXIC SUBSTANCES Several kinds of toxic agents can gener- ate structural and behavioral alterations in animals and humans that result from fetal and perinatal exposure. Experimen- tal studies of time and dose effects should be done with these agents, as has been done for irradiation. Hicks et al. (1961), Hicks and D'Amato (1963), and Berry and Eayrs (1966) showed that one effect of irradiation during fetal life was the alteration of migration pat- terns in the layers of the cerebral cortex. A similar effect was observed in the mouse fetus as a result of subjecting the dams protocols as were the animals mentioned to hypervitaminosis A (Langman and Welch, above with cerebellar microneuronal hype- 1967~. plasia. They with hippocampal micro- Miller (1986) recently showed that ex- neuronal hypoplasia were found to be ex- posure of pregnant rats to ethanol from tremely hyperactive when tested in the ED 6 to ED 23 produced effects similar to open field (Bayer et al., 1973~. They those seen after irradiation, i.e., a defi- were also extremely active, compared with cit in cortical neurons and an alteration control rats, in the running wheels (Peters and Brunner, 1976~. The irradiated rats displayed other behavioral changes usual- ly associated with hippocampal damage, including disappearance of spontaneous alternation in a T maze and deficits in passive avoidance learning (Bayer et al., 1973~. The rats showed deficits at all ages; the deficits differed in severity with age and between tests. Further stud- ies indicated that, as long as the learning tasks ranged from very easy to moderately difficult for normal rats, the irradiated animals performed as well as normal rats (Altman, 1987~. However, when the tasks were more difficult, the irradiated ani- mals were significantly impaired in tac- tile and visual discrimination, in acqui- sition learning, and in reversal learning. Thus, selective microneuronal hypoplas- ia in the hippocampus leads to selective behavioral effects. Altman ( 1986) has in their migration patterns. Although not documented, the cytotoxicity of in- gested ethanol on cortical neuroblasts is a possible underlying mechanism of this observation. Similarly, the administration of lead to rats immediately after birth results in altered hippocampal cytodifferentia- tion, including the presence of smaller numbers of granule cells (Petit et al., 1983; Kawamoto et al., 1984) and behavioral changes. Observation of those effects is confounded by the severe effects of lead on brain endothelial cells (Winder et al., 1983), so a specific effect on the granule cell population cannot be ruled out. RELATIONSHIP BETWEEN MINOR PHYSICAL ANOMALIES AND BEHAVIORAL EFFECTS Agents that influence physical develop-
284 ment are likely to alter behavior. The relationship between phenotypic and be- havioral responses to pollutants is clear- ly exemplified in the study of minor physi- cal abnormalities (MPAs) and behavioral pathology. These observations also sug- gest a potentially useful set of markers for CNS dysfunction. The relationship between MPAs and be- havioral aberrance was first observed in schizophrenics (Waldrop and Halverson, 1972; Goldfarb and Botstein, in press). The following were observed in increased proportion in schizophrenics: excessive- ly fine hair that stands on end, multiple hair whorls, excessively large or small head circumference, epicanthal folds, hypertelorism. low-set ears. adherent earlobes, high arched palate, curved fifth finger, simian palmer crease, spaced toes, and partial syndactyly. That most of those changes are primarily ectodermal in origin suggests that the timing and pathogenesis of the events were shared by alterations in the CNS. The study of MPAs was extended to other behavioral aberrations. In a study of normal 2.5-year-olds, Waldrop et al. (1968) found that, because the number of MPAs was significantly correlated with restless, aggressive impulsive behavior, the MPAs might have been indicators of hyperactivity. Behavior was stable when the subjects were followed up to the age of 5 years. The number of MPAs was found to be negatively correlated with verbal IQ (Rosenberg and Weller, 1973), with full- scale IQ (Waldrop and Halverson, 1972; Firestone and Prabhv, 1983), and with aca- demic achievement (Halvorsen and Victor, 1976~. An apparent sexual dimorphism in the relationship between MPAs and behavior has been observed. Boys with high MPA scores tend to be hyperactive, and girls with high MPA scores seem to display more inhibited, intractable behavior (Waldron , NEURODE~ELOPMENTAL TOXICOLOGY MPAs are increased in autistic children (Steg and Rapoport, 1974~. Quinn and Rapo- port (1974) found an association between increased MPAs and aggression and hyper- activity, but not anxiety, in boys. In the same sample, dopamine p-hydroxylase ac- tivity in blood was correlated with MPA score. MPA scores were higher in hyper- active and retarded boys (Rapoport et al., 1974) and in siblings who were considered mentally normal. Offspring with high MPA scores are more likely to have been the products of complicated pregnancies (e.g., with toxe- mia or prematurity) than of uncomplicated pregnancies (Simonds and Aston, 1981~. They are also more likely to have siblings and parents with high MPA scores; that suggests that both genetic and nongenetic mechanisms were involved in the pathogene- sis (Smalley et al., 1988~. Those observations link a class of rela- tively easily identified and measurable changes in physical structure with abnor- mal CNS development with behavioral defi- cits. Both the physical structures and the CNS are ectodermal in origin. Thus, the two classes of tissue might have re- sponded in an analogous way to a given nox- ious agent. The inference to be drawn is that each class of observations (physical and behavioral) could serve as a marker of the other. MPAs in newborns have also been found in dose-dependent relationship with umbilical cord blood lead concentra- tions (Needleman et al., 1984~. NEUROCHEMICAL EFFECTS For the assessment of nervous system status with chemical methods, samples of various bodily fluids are taken and constituent materials are analyzed (see Table 25-1~. The methods have required the development of high-performance as- says, because the amounts of sample et al., 1976; O'Donnell and Van Tuinan, usually available and the concentrations 1979~. Quinn et al. (1977) classified present in the relevant compartments are infants according to MPA numberintolow, small. The fluids are blood, urine, and middle, and high groups. At 2 years of age, cerebrospinal fluid (CSF). Blood and urine high-MPA boys were more irritable and had are of less utility, because of their re- a higher incidence of night awakening, moteness from the nervous system and the and high-MPA girls were less active and contribution of nonneuronal sources to more withdrawn. the amounts of most substances in these
RESPONSES TO TOX[CAGE[JTS compartments. CSF is not routinely avail- able, inasmuch as its sampling requires medical oversight and involves danger. Advanced techniques of imaging have re- cently enlarged the ways in which biochemi- cal reactions and events in the brain can be measured; these are discussed in the Chapter 30, because they present the most important new opportunities for obtaining markers of neuropsychiatric function. Neurochemical characteristics have been studied in only two major intoxication states: lead poisoning and brain damage induced by ~V-methyl-4-phenyltetrahydro- pyridine (MPTP). The neurochemistry of lead poisoning has been extensively stud- ied in animals (Silbergeld and Hruska, 1980; Winder et al., 1983~. Only recently have attempts been made to extrapolate from the results of those studies to the development of biologic markers in humans. Silbergeld and Chisolm ( 1976) studied monoamine metabolites in urine of lead- exposed children. As shown in Figure 26- 1, there is a correlation between blood lead content and 24-hour urinary excretion of the dopamine metabolite homovanillic acid (HVA) in those children. HVA was meas- ured before initiation of chelation thera- py, within a week after the children were removed from lead-contaminated environ- ments (in all cases, lead paint). Over the long term, urinary HVA content was reduced, 18 16 - . _ 14 - 10 - - o - E 6 > . . 4 ~ 285 as was blood lead content, although both blood lead and urinary HVA remained higher in treated lead-exposed children than in age-matched controls. Other neurochemi- cals reported to be altered by lead in ani- mal models - s uch as GAB A and enkephalins are less available to clinical measure- ment, because they require CSF, and have not been investigated in humans. MPTP is a contaminant in some Designer drugs or meperidine derivatives with opiatelike characteristics. After the remarkable finding that some addicts had acute-onset necrologic disorders that were indistinguishable from so-called idiopathic Parkinson's disease (Langston et al., 1983), attention focused on MPTP as the active pathologic agent. MPTP was found to be a specific basal ganglia toxin that damages the same nigrostriatal dopa- minergic pathways that are affected in parkinsonism (Kopin and Markey, 1988~. The mechanism of action of MPTP involves uptake into dopaminergic neurons, inter- action with oxidases within the neurons, and selective cell killing, possibly by the generation of free-radical oxygen or hydroxyl radicals. As a dopaminergic tox- in, MPTP could be expected to reduce output of dopamine metabolites from brain into CSF; this has been demonstrated in primate models of intoxication (Kopin and Markey, 1988~. . r = .729 p <.001 · / 1.0 1.5 2 0 2 5 3 0 3 5 4 0 PbB(pM/L) FIGURE 26 1 Correlation between blood lead and 24 hour urinary excretion in children. Source: Silbergeld and Chisolm, 1976.
286 BEHAVIORAL EFFECTS Behavior is the observable response of an organism to changes in the external or internal environment. That definition includes actions ranging from reflex re- sponses to the solving of complex problems and performance on psychometric tests and in social situations. Behavioral analysis can proceed at any of those levels. The investigator of neurobehavioral function is confronted with a choice be- tween attempting to characterize a pos- sible deficit by measuring a single, iso- lated function and scaling more complex integrative behaviors. The choice might present a trade-off between the precision of a test that measures a single function (or as close to a single function as pos- sible) and the relative imprecision of a test instrument that measures the sub- ject's integrative capacity in a more com- plex single-demand task or in a number of tasks. The use of measures of single functions can be more effective in detect- ing neurotoxic deficits than gross-per- formance tests, because they are more fo- cused and demanding (Smith, 1985~. Tests of integrative capacity might yield in- creased sensitivity, but lead to less pre- cision as to the location and degree of a deficit. A number of outcomes are presented below, roughly in increasing order of functional complexity. In each instance, their utili- ty when testing for lead exposures is no- ted, not because that represents their sole utility, but because the effects of lead exposure is the most widely studied toxicity in humans. · · Nerve conduction time. Measurements of nerve conduction time have been useful in assessing neurotoxicants that act on myelin development or on Schwann cells. Occupational exposures to lead at doses that did not produce symptoms were associ- ated with impairments in nerve conduction (Seppalainen and Hernberg, 1980~. Chil- dren exposed to lead from industrial sources were found to have dose-dependent nerve-conduction slowing (Landrigan et al., 1976~. Because normal variation is large, assessment of nerve conduction time NEURODE~ELOPMENTAL TOXICOLOGY is not well suited as a screening device in neurotoxicant exposure. · Sensory psychophysical functions and scotopic vision. Measures of visual acuity in bright light were not related to lead exposure in primates. Visual acuity in the dark was diminished in primates exposed to lead (Bushnell et al., 1977~. The re- sults suggest specific impairment of rod function. · Evoked potentials. Because evoked po- tentials recorded from the scalp or spinal cord reflect activity in multisyn- aptic pathways, evoked potentials can provide useful information in assessing sensory transmission from the periphery to the cerebral cortex. Three types of evoked potentials have been recorded: auditory evoked potentials for which the stimuli are generally pure tones or clicks, but can be phonemes or words; visual evoked potentials for which the stimuli are stro- boscopic light flashes or checkerboard patterns on computer monitors; and somato- sensory evoked potentials for which the stimuli are brief electric impulses deliv- ered to the skin. Auditory evoked poten- tials have been found to be altered in ex- posure to lead (Otto et al., 1981), carbon monoxide (Groll-Knoff et al., 1978), and trichloroethylene (Winneke et al., 1978~. Visual evoked potentials have been found to be affected in exposure to xylene (Seppalainen et al., 1981), methyl mercury (Iwata, 1980), and n-hexane (Seppalainen et al., 1979~. Somatosensory evoked poten- tials have been reported to be altered in exposure to lead (Seppalainen, 1978), but not in exposure to n-hexane severe enough to be symptomatic (Zappoli et al., 1978~. · Auditory discrimination. Auditorydis- crimination can be tested against various masking backgrounds, such as taped sounds containing a signal against background noises of increasing loudness (e.g., elec- tric fan sounds or noise from a cafe) (Gold- man et al., 1970~. In studies of asympto- matic children, those with the higher con- centrations of lead in their teeth had lower scores on this measure (Needleman etal.,1979~.
RESPONSES TO TOXIC AGENTS · Vibration sense. Vibration sense can be measured with a number of methods; the most simple (and least sensitive) uses a tuning fork. Computerized methods with increased sensitivity and precision are available (Maurissen and Weiss, 1980~. Decreased vibratory sensitivity is ob- served in many conditions that involve the central and peripheral nervous sys- tems. Among them are systemic disease, such as diabetes, chronic liver failure, pernicious anemia, peripheral neuropath- ies, syphilis, spinal-cord lesions, and uremia; exposure to pharmaceuticals, such as isoniazid, phenytoin, vincristine, and glutethimide; and exposure to chemi- cals, such as acrylamide, arsenic, n-hex- ane, mercury, and methylbutylketone (Maurissen, 1985~. · Motor function. Quantitative measures of tremor have been evaluated in methyl mercury exposure. Quantitatively in- creased tremors were shown when clinical studies were noninformative. Studies of motor patterns of children classified as to lead exposure are under way. One study, which measured on-task behavior of chil- dren differentially exposed to lead, found that high-lead children spent more time in off-task behavior in the classroom- e.g., out-of-chair activity, staring out of the window, and talking to classmates (Needleman and Bellinger, 1981~. · Attention. Attention is a complex func- tion of arousal, vigilance, and resistance to distraction. A number of measures of it are available. In the Continuous Per- formance Test, a measure of vigilance (Rosvold et al., 1956), the subject is presented with a set of stimuli (letters of the alphabet) on a screen for short in- tervals in rapid succession. The critical stimulus (the letter X) is presented at a predetermined probability. The demand task is to press the response key when the critical stimulus appears. The number ~ ~ · e · - ot errors ot omission, commission, and latency to response can be determined. Reaction time with various intervals of delay has been used to discriminate among children who differed as to lead exposure 287 (Needleman et al., 1979; Yule et al., 1981; Hunter et al., 1985~. · Visual-motor integration. Measures of visual-motor integration, such as the Bender-Gestalt test (Trillingsgaard et al., 1985), have found wide application in the study of brain damage in children. A number of skills are called on: spatial v~sua~za~on, eye - hand coordination, and visual-spatial memory. · Speech and language function. Speech and language function is the sum of many competences and can be perturbed at many levels. At the perceptual level, auditory acuity for pure tones and the ability to screen out background distractions can be measured (Goldman et al., 1970~. Short- term memory and the ability to discriminate patterns are testable (Seashore Rhythm Test). The ability to comprehend language is testable with many instruments, such as the Token Test, the verbal subtests of the WISC-R, and the Illinois Test of Psy- cholinguistic Abilities. · Psychometric intelligence. Studies of psychometric intelligence in children have been widely used in recent years to study lead toxicity, exposure to polybro- minated biphenyls (PBBs), and fetal al- cohol exposure. Three tests have been used generally: the Bayley Scales of Infant Development for children between 6 months and 3 years old, the McCarthy Scales for children more than 3 years old but less than 5 years old, and the Wechsler Intelligence Scales revised for children over 5 years old. The most sensitive subscales of these instruments appear to be the verbal and general cognitive index. Numerous studies of effects of lead exposure at low dose in children have shown IQ deficits after con- trol of relevant covariates (U.S. EPA, 1986~. · Social behavior. Attention to cogni- tive, perceptual, and motor competences should not direct attention away from so- cial behavior of children. Three groups of investigators have rated classroom behavior in lead-exposed children with structured questionnaires (Needleman et al., 1979; Yule et al., 1981; Hatzakis
288 et al., 1987~. Teachers blind to a sub ject's lead exposure reported a dose dependent increase in nonadaptive class- room behavior, such as distractibility, - NEURODEKELOPMENTAL TOXICOLOGY inability to work independently, dis- organization, hyperactivity, impulsivi- ty, and inability to follow directions (Needleman et al., 1979~.
