Parkinsonism (P) with progressive dementia (D) of the Alzheimer type is recognized as a clinical variant of a form of amyotrophic lateral sclerosis (ALS) that has occurred in high incidence among the Chamorro people of the islands of Guam and Rota in the Marianas chain of Micronesia. The declining annual incidence, upward shifting of the age of onset, narrowing of the sex ratio, and occurrence of the disease among non-Chamorros, point to a disappearing environmental causation peculiar to the traditional culture of these islands. Evidence is presented in support of the proposal that heavy use of certain toxic plants, notably cycads, a traditional source of food and medicine for the Chamorro people, plays an important etiological role. Clinical and epidemiological approaches are offered to test for a relationship between ALS/P-D and long-latency plant toxicity.

For more than 150 years, Chamorro natives of the Mariana Islands in the Western Pacific Ocean, have developed fatal paralysis in middle and later life, which we term amyotrophic lateral sclerosis/parkinsonism-dementia (ALS/PD). The cause of the disease might be exposure to seeds of the indigenous cycad. Motor system disease is induced in cynomolgus monkeys by feeding them beta-N-methylamino-L-alanine (BMAA), an amino acid present in cycad seeds. We believe that the cycad seeds which usually cause no immediate adverse symptoms when prepared and eaten as flour, or applied topically as medicine, can give rise to widespread and severe nerve cell degeneration after a latency of many decades. Furthermore, it may be that only a single exposure to this potent but silent toxin(s) can result in fatal neurological disease years later.

Neurotransmitters of the basal ganglia are of three types: I, amino acids; II, amines; and III, peptides. The amino acids generally act ionotropically while the amines and peptides generally act metabotropically. There are many examples of neurotransmitter coexistence in basal ganglia neurons. Diseases of the basal ganglia are characterized by selective neuronal degeneration. Lesions of the caudate, putamen, subthalamus and substantia nigra pars compacta occur, respectively, in chorea, dystonia, hemiballismus and parkinsonism. The differing signs and symptoms of these diseases constitute strong evidence of the functions of these various nuclei. Basal ganglia diseases can be of genetic origin, as in Huntington's chorea and Wilson's disease, of infectious origin as in Sydenham's chorea and postencephalitic parkinsonism, or of toxic origin as in MPTP poisoning. Regardless of the etiology, the pathogenesis is often regionally concentrated for reasons that are poorly understood. From studies on Parkinson and Huntington disease brains, evidence is presented that a common feature may be the expression of HLA-DR antigen on reactive microglia in the region where pathological neuronal dropout is occurring.

This paper reviews current concepts concerning the anatomical organization and functions of the basal ganglia focusing mainly on the role of the basal ganglia in control of movement. Results of several studies which have been carried out on patients with Parkinson's disease are also reviewed and discussed as examples of altered normal physiological mechanisms.

Disruption of neural activity within the basal ganglia of experimental animals causes selective learning deficits in tasks requiring switching between response strategies. These data along with reports of both general and specific intellectual impairment in patients with neurodegenerative disorders such as Parkinson's disease, appear to support the theory of cognitive functions of the basal ganglia. Recent studies have failed to confirm general cognitive or memory deficits in parkinsonian patients, but have identified deficiencies in devising and executing certain cognitive strategies. Following the lead of theorists such as Squire and Mishkin, this brief review emphasizes the distinction between procedural and declarative knowledge and examines the possible role of the basal ganglia in the acquisition and retention of procedural knowledge.

The major input and output pathways of the mammalian striatum have been well established. Recent studies have identified a number of neurotransmitters used by these pathways as well as by striatal interneurons, and have begun to unravel their synaptic connections. The major output neurons have been identified as medium spiny neurons which contain ɣ-aminobutyric acid (GABA), endogeneous opioids, and substance P. These neurons project to the pallidum and substantia nigra in a topographic and probably chemically organized manner. The major striatal afferents from the cerebral cortex, thalamus, and substantia nigra terminate, at least in part, on these striatal projection neurons. Striatal interneurons contain acetylcholine, GABA, and somatostatin plus neuropeptide Y, and appear to synapse on striatal projection neurons. In recent years, much activity has been directed to the neurochemical and hodological heterogeneities which occur at a macroscopic level in the striatum. This has led to the concept of a patch-matrix organization in the striatum.

The study of neurotransmitter receptors aids in the understanding of the normal anatomy, pharmacology, therapeutics and pathophysiology of disease processes involving the basal ganglia. Receptors may be studied in vitro by homogenate binding experiments, enzyme analysis or quantitative autoradiography and in vivo with positron emission tomography. In the substantia nigra (SN), receptors have been identified for somatostatin, neurotensin, substance P, glycine, benzodiazepine and GABA, opiates, dopamine, angiotensin converting enzyme (ACE) and serotonin. The striatum has receptors for dopamine, GABA and benzodiazepines, acetylcholine, opiates, substance P, glutamate and cholecystokinin. GABA and benzodiazepine receptors are also located in the globus pallidus. In Parkinson's disease, striatal dopamine D-2 receptors are elevated in patients that have not received L-DOPA therapy. This supersensitivity is reversed with agonist therapy. Muscarinic binding to cholinergic receptors seems to correlate with dopamine receptors. Delta opiate receptors are increased in the caudate and mu binding is reduced in the striatum. In the SN of patients with Parkinson's disease, there is reduced binding of somatostatin, neurotensin, mu and kappa opiates, benzodiazepine and GABA and glycine. In Huntington's disease, there is reduced binding of GABA and benzodiazepines, dopamine, acetylcholine, glutamate and CCK. There is increased binding of GABA in both the SN and globus pallidus. Glycine binding is increased in the substantia nigra and ACE is reduced.

The existence of subtypes of dopamine receptors as defined by Kebabian and Calne is now well accepted. The Dl receptor is typically associated with stimulation of adenylate cyclase, while the D2 receptor is either independent of adenylate cyclase or mediates its inhibition. Considerable interest has been generated by the potential physiological and clinical roles of these receptor subtypes. Availability of agonists and antagonists specifically acting at the D1 or D2 receptor site has stimulated research to characterize the functional effects of each receptor subtype. This might facilitate the development of effective compounds to control the signs and symptoms of Parkinson's disease and perhaps prevent the induction of debilitating side effects. Recent evidence indicates that Dl receptor stimulation is required to obtain full expression of the D2 receptor site, which has typically been associated with the clinical benefits of dopaminergic therapy. Both pre- and postsynaptic location of the receptors must also be taken into consideration as well as involvement of other neuronal systems. It appears that in PD, progressive involvement of the dopaminergic pathways is the principal pathological course, however, noradrenergic and serotonergic pathways are likewise involved. This multineuronal involvement suggests that drugs acting specifically at receptor sites, located on both pre- and postsynaptic neurons, might be required at different times during the course of the disease process to control its symptoms and/or the complications occurring after long-term treatment with a given drug.

Descriptive data from several studies suggest variations in the frequency of Parkinson's disease in different population groups. Door-to-door surveys were carried out among a biracial U.S. population (blacks and whites) and in communities in Nigeria and the People's Republic of China. The U.S. investigation revealed no substantial differences in the age-adjusted prevalence ratios by race. However, blacks in Nigeria have a much lower prevalence ratio than blacks in the U. S., suggesting an environmental etiologic factor. Prevalence ratios derived from China are also lower than the U.S. figures. Studies of temporal trends in the incidence rates in one U.S. population (Rochester, Minnesota) show virtually no change over 35 years, indicating that the primary cause(s) of Parkinson's disease must have been present in this nonindustrialized community for many years. Analytic studies generally reveal an inverse association between Parkinson's disease and cigarette smoking, although epidemiologic evidence does not support a direct protective effect of smoking. Preliminary investigations suggest an increased risk associated with the rural environment and the consumption of well water. Further studies are required to discover as yet unknown environmental factors that heighten the risk of Parkinson's disease.

In 1984 we made the first observation of a correlation between early age exposure to rural environment (and drinking well water) and development of idiopathic Parkinson's disease (IPD).13 These findings were subsequently confirmed elsewhere (Barbeau, 1985 ;25 Tanner, 198526). Analysis of all early age onset IPD (EPD) cases born and raised in Saskatchewan revealed that 20 of 22 had exclusively rural exposure during the first 15 years of life. This distribution was significantly different from the general population (p = 0.0141). Further study of the EPD group included sampling and metal analysis of childhood sources of drinking water in 18 cases and 36 age and sex-matched controls. Water collected from the two groups was analyzed for 23 metals (including 7 elements implicated in the etiology of IPD). There was no difference in the metal composition of the water between the two groups.17 Finally, a review of herbicide and pesticide use in Saskatchewan agriculture was undertaken to determine if there was an increased incidence of EPD following utilization of any particular chemical. No increase was found in the incidence of EPD with the introduction of any pesticide or herbicide, including Paraquat, for agricultural use. We conclude that there is a strong correlation between early age rural environmental exposure and development of IPD. We believe well water is a likely vehicle for the causal agent, but neither water metal concentration nor any of the herbicides and pesticides used in Saskatchewan agriculture are related to the cause.

Parkinson's disease (PD) has been proposed to result from the interaction of aging and environment in susceptible individuals. Defective metabolism of debrisoquine, inherited as an autosomal recessive, has been associated with this susceptibility. In 35 PD patients and 19 age-matched controls, no significant differences in debrisoquine metabolism were found, although a trend to impaired metabolism was noted in patients with disease onset ≤40. Foci of PD patients were associated with rural living and well water drinking, or rural living coupled with market gardening or wood pulp mills. In a questionnaire survey, patients with PD onset ≤age 47 were significantly more likely to have lived in rural areas and to have drunk well water than those with onset ≥age 54 (p≤0.01). Because of population mobility in North America, a case-control study designed to test environmental, occupational, dietary and other proposed risk factors for PD was conducted in China, where the population is more stationary and the environment more stable. No significant differences in incidences of head trauma, smoking or childhood measles were found between patients and controls.

Progressive degeneration of functionally related groups of neurons occurs in certain infective, toxic, nutritional and genetically determined neurological diseases. It also takes place in normal aging, and several of the regions that undergo selective decay with the passage of time seem to be the same target regions that are afflicted in degenerative disorders such as Parkinson's disease, Alzheimer's disease and amyotrophic lateral sclerosis (ALS). Infective etiology is relatively easy to exclude by a combination of immunological tests and transfer experiments. Genetic causation can be rendered unlikely when large kindreds are available for study. Nutritional deprivation and acute or subacute toxicity are accessible to explanation by examining the environment. The most difficult mechanism of pathogenesis to refute is chronic toxic damage, where the lesion may derive from long-term exposure to a relatively widespread noxious agent or agents. Variations in involvement of individuals within a population may stem from differing capacities to activate or inactivate a toxin. Inherent in this concept of etiology is recognition that compensatory potential within the central nervous system may contribute to prolonged existence of subclinical lesions so that a latent period may exist for several decades, between causal event and the onset of symptoms. Furthermore, progressive clinical deterioration may take place even though the cause may have been transient, many years before. The histological features associated with Parkinson's disease, Alzheimer's disease and ALS may be nonspecific indicators of neuronal “illness”, there being a predilection for certain morphological markers to appear more frequently in particular circumstances and particular regions associated with the pathology of particular diseases.

MPTP induces parkinsonism in monkeys by destruction of the substantia nigra, pars compacta. It can also damage ventral tegmental dopamine neurones and the noradrenergic locus coeruleus, both of which may be affected in Parkinson's disease. Motor symptoms in MPTP-treated monkeys respond readily to levodopa or dopamine agonist therapy. Administration of levodopa over 4-8 weeks leads to the emergence of “peak- dose” dyskinesia. Such abnormal movements are not seen following challenge doses of levodopa in animals not on long-term therapy. Radioligand studies reveal a 40-180% increase in D2 receptor binding in the striatum of parkinsonian monkeys. 2-deoxyglucose studies of regional brain metabolism indicate that MPTP-induced parkinsonism is characterised by abnormally increased activity of medial pallidal neurones which project to the thalamus and pedunculopontine nucleus and reduced activity of subthalamic nucleus neurones.

The MPTP-treated mouse has proven to be a valuable model of parkinsonism. For example, C57 black mice treated with MPTP exhibit a large decrement in the neostriatal content of dopamine and its metabolites, a marked reduction in the capacity of neostriatal synaptosomal preparations to accumulate [3H]dopamine, a large decrease in neostriatal tyrosine hydroxylase activity, a marked loss of nerve cells in the zona compacta of the substantia nigra, and pronounced behavioral deficits. These biochemical, pathological and behavioral deficits are similarly observed in MPTP-treated primates and in humans with idiopathic parkinsonism. A great deal of our current knowledge concerning MPTP has come from experimentation carried out in the mouse.

Glutaric aciduria type I is an autosomal recessive metabolic disorder of children associated with severe dystonic motor disturbances and degeneration in the cerebral cortex, striatum and cerebellum. Biochemical studies demonstrate a deficiency in the enzyme glutaryl-CoA dehydrogenase. This enzyme metabolizes substrate derived from dietary tryptophan that could otherwise be converted to quinolinic acid within the brain. The law of mass action predicts that the production of quinolinic acid should be increased in glutaric aciduria type I. Quinolinic acid is a potent neurotoxin and convulsant when it is injected into the central nervous system of experimental animals. This paper argues that quinolinic acid may accumulate within the brain and cause the neuropathology of glutaric aciduria type I.

The local concentration of 6-[18F]fluoro-L-dopa (18F) reflects the activity of aromatic acid decarboxylase (AADC), the enzyme that generates dopamine from its precursor amino acid, L-dopa. In young healthy adults, the local concentration of 18F, and hence AADC activity, is constant in coronal slices taken in a rostrocaudal direction. With increasing age a gradient representing decreasing activity in the putamen develops. This decrease is less marked than was expected from the literature. In five children with primary dystonia, the striatal distribution of 18F resembled that seen in the normal older adults. In established clinical Parkinson's disease the rostrocaudal gradient becomes steep; the putamen is more damaged.

The development of positron emission tomography (PET) has enabled us to perform in vivo measurements of certain aspects of regional cerebral function. Regional cerebral glucose metabolism may be readily quantified with [18F] fluoro-2-deoxyglucose (FDG) and presynaptic dopaminergic function may be studied with the labelled dopa analog 6-[18F] fluoro-L-dopa. We have applied a model to the analysis of 6-FD/PET data with which in vivo age-related changes in dopaminergic function may be demonstrated in normal subjects. With this technique, we have studied a series of asymptomatic MPTP-exposed subjects and have shown evidence of subclinical nigrostriatal pathway damage. Studies of regional cerebral glucose metabolism with FDG in early Huntington's disease have shown a characteristic impairment in caudate function which precedes the development of caudate atrophy. In addition, some asymptomatic individuals who are at risk for HD have caudate hypometabolism. We feel that, at the present time, PET provides information which is complementary to the clinical examination in establishing a diagnosis of HD. In the future these studies may also help in the investigation of at risk individuals

Levodopa is currently used at all stages of Parkinson's disease, particularly from Stage 3 onward. Most patients start levodopa within four years of the onset of disease, earlier in the akineto-rigid patients (average delay of 2.1 years) than in those where resting tremor predominates (average delay of 3.4 years). Advanced age (>80) is no deterrent to the use of levodopa if required. Wearing off is most noticeable in Stage 3 or after 10 years of therapy with levodopa. An inverse correlation was found between age and wearing off, suggesting a marked sensitivity of younger patients to levodopa. Future studies on the use of dopamine agonists should thus be stratified according to age.

As the direct agonist with the widest clinical use, bromocriptine provides a unique window into the clinical spectrum of Parkinson's disease. The efficacy of bromocriptine for therapy of de novo Parkinson's disease has recently been confirmed using a double-blind design with L-Dopa (Sinemet). Over a period of 5.5 months, bromocriptine was found to be as effective as L-Dopa in reducing the functional and neurological disability of Parkinson's disease. This study complements others and demonstrates a role for bromocriptine as de novo therapy. A longitudinal study comparing bromocriptine with L-Dopa is underway, but previous observations with bromocriptine suggest modest, transient beneficial effects with significantly less fluctuation of disability and less dyskinesia when used alone or in combination with L-Dopa. The transient benefits of bromocriptine on progressive disability suggest that both pre-and post-synaptic defects are eventually involved in Parkinson's disease. While agonists with improved efficacy and minimal side effects are required for symptomatic treatment of Parkinson's disease, strategies to protect pre- and post-synaptic neuron populations against progressive dysfunction must be developed.