Alzheimer's Disease
Alzheimer's disease (AD) is the most prevalent form of senile dementia in the United States. As the American population ages, AD's social and financial costs escalate. Unfortunately, our poor understanding of AD etiology and neuropathology has impeded therapeutic progress. Several animal models have been extensively studied in an effort to improve our understanding and treatment of AD.
Background: Alzheimer's Disease
Alzheimer's disease involves progressive cerebral atrophy, causing a gradual decline in cognitive function. The insidious course begins with forgetfulness which is often attributed to normal aging. After several years, patients exhibit confusion, personality changes, and language and attention difficulties. Physical health and level of arousal, however, are not significantly affected in AD. The course of the disease is about ten years, and it reduces life expectancy at the time of diagnosis by 1/2 to 2/3.(1) Death is often due to pneumonia or other secondary diseases.(2)
AD is associated with specific cognitive and personality changes. The earliest symptoms are generally related to short-term memory. Later symptoms include poor social functioning and difficulties performing routine daily tasks. Finally, patients are unable to orient themselves with regard to to time or place, and they do not recognize other individuals, including members of their own family. Problems with vision, coordination, and speech develop in the disease's later stages.
Grossly, AD patients have extensive cortical, hippocampal, and amygdalal degeneration. Microscopically, neuron and dendritic spines decrease in number, and amyloid-containing neuritic plaques and neurofibrillary tangles develop. The causes of these pathological changes are poorly understood. One theory is that free oxygen radicals kill conical cells, leading to arnyloid degradation products.(3) It has been well documented that older cells have a limited capacity to reduce free oxygen radicals.(3)
In addition to the characteristic distribution of neuronal damage, there are distinctive changes in neurotransmitter levels. The causes and clinical relevance of neurotransmitter abnormalities are subjects of considerable speculation and confusion.(4-6) One widely held hypothesis is that reduced choline acetyltransferase, which leads to reduced acetylcholine, causes pathology or death of many acetylcholine-producing neurons, The cholinergic system, however, interacts widely with other neurotransmitter systems. Consequently, it is not known if the reduced serotonin and noradrenaline levels seen in AD are secondary to changes in acetyicholine-producing neurons or if they are due to other neurochemical disturbances.
AD pathogenesis is not well understood. Amyloid protein deposits may precipitate the degenerative process.(7) However, the amyloid deposits may be secondary to neuronal death or may appear coincidentally in time with the degeneration. Another theory is that an excess endogenous neurotoxin plays an important role. Some investigators, recognizing that aluminum is found in the neurofibrillary tangles, have suggested aluminum as a causative factor. There is evidence, however that aluminum is not an etiological agent, but is sequestered and trapped by the tangles as they form.(3) Perhaps trauma causes neuronal death and precipitates the disease process in some patients. Finally, epidemiological studies demonstrate that relatives of AD patients have an increased risk of developing the disease, indicating that genetics is a likely contributing factor in many cases.(8)
Background: Animal Models
Researchers have attempted to gain insights into AD from a wide range of animal studies. While acknowledging that no animal model faithfully reproduces AD, many suggest that refined animal models may someday enhance our understanding and treatment of the disease. This report examines the probable utility of the most prevalent animal models of AD.
Several models involve impairment of the cholinergic system of young animals. The three most widely used methods are destruction of the nucleus basalis of Meynert (NBM), a major nucleus of the cholinergic system, intracerebroventricular administration of ethyicholine mustard aziridinium ion (AF64A, a selective cholinotoxin), and scopolamine admininstration.(9) Another common type of research uses older animals to study age-related cognitive changes. Most of the research involving manipulations of the cholinergic system has used rats; most of the research with older animals has used rhesus and squirrel monkeys.
Part 1: Comparison of Animal Models to Human Alzheimer's Disease
Cholinergic Models Clinical Presentation
The three leading cholinergic models selectively alter the brain's acetylcholine status without extensively changing the concentrations of other neurotransmitters. While some of the choline level alterations are similar to those in AD, there are significant differences. For example, an important feature of AD is cholinergic disturbance in the hippocampus, but this stucture is spared in NBM-lesioned rats.(7,9) At the synaptic level, lesioned rats have a decreased affinity for acetylcholine, that is also seen in AD. However, the rats also have a reduction in receptor number that does not occur in AD. As in AD, there is a sensitivity to scopolamine. Unlike AD patients, however, scopolamine-treated and NBM-lesioned rats respond dramatically to physostigmine, a drug that impedes acetylcholine breakdown. The effect of physostigmine after AF6A administration is unknown.(9)
The models differ from AD in many respects, perhaps because these models manipulate only one system and AD appears to involve several neurochemical alterations. For example, the NBM-lesioned and scopolamine-treated rats do not reproduce the characteristic profile of glucose metabolism. AD patients, unlike the lesioned rats, have reduced glucose consumption in the temporoparietal cortex with sparing of areas to which these lobes project.(9) Unlike AD patients, NBM-lesioned animals have normal levels of somatostatin and norepinephrine.(4,9) Interestingly, in some of the rat models, the hippocampus and frontoparietal cortex show "tangle-like" structures as well as cell loss. However, unlike AD tangles, these tangles do not stain for amyloid.(9) Finally, NBM-lesioned rats exhibit gross motor and behavioral changes not observed in AD, including motor incoordination and hyperactivity. Also, unlike AD patients, these rats generally do not eat or drink at all.(9)
The cholinergic manipulations in rats impair memory and spatial localization, which are also impaired in AD patients. It is not clear, however, whether gross tests of cognitive function in laboratory rats are relevant to the specific cognitive impairments found in AD. For unknown reasons, NBM- and AF6A-lesioned rats show recovery of cognitive and behavioral alternations, despite general persistence of cholinergic deficits.(9) In NBM-lesioned animals, Smith notes, "Reversal of behavioral deficits with pharmacological agents and recovery over time has been observed, in the presence of persistant cholinergic deficit."(8) Thus, the disease's natural history in these animals differs from AD's relentless, progressive course.
A limitation of these animal models, Coyle et al. recognize, is their failure to replicate the complexity of the human central nervous system: "Since more than one neurochemical system obviously plays a role in memory, perhaps models of human disease should involve manipulation of more than one system at a time."(4) Similarly, Smith has observed, "Current animal models of AD are limited by the behavioral and neurophysiological findings that have been reported in AD and the poor understanding of the interrelation among these factors."(9) Finally, Wenk has commented on another difficulty using the animal models -- the lack of sufficient lesion specificity:
Use of these pharmacological agents does not provide information into the regional specificity of the processes and the brain areas involved. Scopolamine injections impair cholinergic function in the thalamus, hippocampus, caudate, neocortex, and any other area that has muscarinic cholinergic receptors. . . . Ibotenic acid (one of the agents used in NBM lesions) is not selective for a particular neurotransmitter system.(10)
Pathogenesis
Cholinergic models of AD are based on the cholinergic hypothesis. In the early 1960s, Whitehouse noted effects of acetyicholine and its analogues on learning.(11) Similarly, Drachman has shown, in young volunteers, that scopolamine causes transient memory impairment similar to the memory dysfunction seen in elderly people.(12) Further support for the cholinergic hypothesis comes from evidence that AD patients suffer from lowered choline levels. Lesioning of the rat NBM significantly reduces choline acetyltransferase concentration in the cerebral cortex. This lesioning is done with many selective and non-selective neurotoxins, including quinolinic acid (an endogenous neurotransmitter that increases in concentration with age in humans), ibotenic acid, and kainic acid; AF64A, a chemical that poisons the acetyicholine system; and scopolamine, a reversible muscarinic receptor antagonist.(9) While such cholinergic models tend to reproduce grossly some of the choline alterations found in AD, it is highly unlikely that these artificial laboratory manipulations resemble the causative insult(s) of AD.
In conclusion, there are important differences in the etiopathology, symptomatology, and neurochemistry between the cholinergic models and human AD. This restricts the models' ability to provide insight into AD's nature and suggest effective treatments,
2) Aged Animals
Clinical Presentation
Aged primates exhibit certain similarities to AD patients. While these primates do not have neurofibrillary tangles, they do have plaques in neural tissue comprised of amyloid surrounded by abnormal neurites and reactive glial cells. As in AD, these plaques are located in the cortex, hippocampus, and amygdala.(13) Plaques are also seen in normal elderly people, however, and their contribution to AD symptoms is unclear. Adult monkeys have a decreased number of receptors in certain brain regions. AD patients, in contrast, do not have much reduction in receptor number as revealed by autopsy.(14)
Investigators recognize that aged primates are not bonified models of AD. The primates are used mainly to study age-related changes in cognitive ability. Aged primates perform poorly on several tests designed to measure cognition.(15,16)
Aged monkeys also perform worse than young monkeys on a test of visuospatial function involving removal of lifesavers from a twisted rod.(16) Many of these general effects are seen in both normal human aging and in AD. However, as Moss et al. have observed, "at present there is little correspondence between the tasks used to assess memory function in aged humans and those used in monkeys."(16) In order to compare the cognitive dysfunction in aged monkeys to the deficits found in AD patients, one must assume that the laboratory variables studied (given generic terms such as "visuospatial difficulties" and "short-term memory") relate meaningfully to the cognitive dysfunction of AD patients. Lister has recognized the difficulty of applying results from the artificial laboratory environment to human patients:
A criticism that has been leveled against many laboratory tests of learning and memory is that they are poor gauges of the performance of the subject (either human or animal) in the real world. Furthermore, the ability to extrapolate across species must also be considered. We are only able to collate findings from experiments using different methods of assessing cognition and extrapolate these findings to humans if we understand the cognitive significance of performance in the tests.(17)
Pathogenesis
Unlike the artificial manipulations of the cholinergic models, aging is a natural process. It is not clear, however, that any non-human animal manifests neurological changes analogous to AD. While it is possible that age-related neurological changes and cognitive deficits in laboratory primates may enhance our understanding of normal aging processes in the brain, the relevance of the non-human primate research for determining AD etiology is questionable. Indeed, Coyle et al. have noted:
Because AD was initially considered to be the consequence of brain aging, much effort has been expended in preclinical research in defining the synaptic neurochemical alterations associated with aging in experimental animals. However, the applicability of these findings to human AD remains dubious because of differences in species and strain responses to aging, the brief duration of life of most experimental animals and the absence of neuropathologic stigmata of AD in sub-primates.(4)
3) New Animal Models
Two rodent "models" of AD have recently been described. Although investigators working with these models express optimism that they will provide fresh insights, their utility remains unproven. Kowall and colleagues injected beta amyloid, a brain protein, into the hippocampus of rats and produced neuron death similar to that of AD. However, Kowall et al. have acknowledge that artifact may contribute to neuron damage: "the neurotoxicity of β amyloid may be accelerated by the rapid presentation of relatively high local concentrations of solubilized β amyloid."(8) Kowall et al. also have reported that substance P. another brain protein, protects against this damage.(18) This effect, which was first demonstrated in vitro, may or may not be relevant to AD. Indeed, co-worker Dr. Bruce Yanker has stated, "We do not by any means claim to have created Alzheimer's disease in the rat."(19)
Wirak et al. have used human genes to induce β amyloid deposition in the hippocampus of transgenic mice.(20) However, the deposits differ from those of AD patients. Whereas the human β amyloid deposits are extracellular, they are intracellular in the mice. Also, the mice have not exhibited neuronal degeneration or signs of CNS dysfunction after one year. Whether AD-like changes are seen after one year remains to be seen.
Part 2: Clinical Review Articles
Three recent review articles address different aspects of AD. Two of these articles, "Alzheimer's disease: Evolving clinical concepts and management strategies"(1) and "The risk factors for Alzheimer's disease: A review and a hypothesis,"(3) contain 15 and 123 references, respectively, yet neither mentions research with animal models of AD. The third, "The neurobiology of Alzheimer's disease" by Henderson and Finch, contains 309 references, only two of which deal with animal models of AD.(7) First, Henderson and Finch note that amyloid plaques occur in aging monkeys as well as AD patients. However, the finding that amyloid plaques are not specific for AD did not require animal studies. This was readily apparent from human autopsies. Henderson and Finch's second reference is that scrapie, a disease primarily of sheep, can also infect humans with a syndrome vaguely similar to AD. Henderson and Finch have written that this "has fueled speculation that this or other viruses might be implicated in AD."(7) They acknowledge,however, "the search for conventional virus in AD brain has been unrevealing."(7) Henderson and Finch have made a final observation regarding MPTP toxicity, which causes a Parkinsonian syndrome in both humans and other animals. They do not identify MPTP toxicity as a model of AD, but they suggest that MPTP studies have illustrated that "environmental toxins can be linked to chronic neurodegenerative illness."(7) Thus, these clinical review articles do not reveal any important contributions from animal models to our understanding or treatment of AD.
Part 3: Clinically Important Questions
Clinicians need answers to several questions in order to improve AD management. First, what is AD's underlying pathological process? An answer to this question would greatly assist therapeutic strategies. Bartus (21) and others have advocated the development of animal models that reproduce AD pathology more closely than those currently available. However, Haroutunian has commented, "Even if an animal model could be developed which mirrored all the known neurochemical and neuropathological aspects of the disease process, it would still only reflect those aspects of AD which have been characterized thus far."(22)
Currently available animal models are compromised by their lack of specificity with regard to neuronal damage and by our poor understanding of both human and animal neurophysiology. However, even if scientists developed a model that faithfully reproduced AD's known symptoms and neuropathology, this would not necessarily enhance our knowledge of human AD. Because AD appears to be a uniquely human condition, it seems unlikely that animal models will increase our understanding of this disorder.
Much animal research has attempted to identify the metabolic and neurochemical derangements of AD. To date, autopsy studies have been of great importance. Most likely, further insights will derive primarily from human clinical research. Many non-invasive procedures on humans permit study of the intact disease process. For example, extensive research on patients and volunteers has investigated possible therapeutic effects of scopolamine, physostigmine, and other drugs that affect the cholinergic system.(9,12,23) Because human subjects can generally perform more sophisticated cognitive tasks, they are preferable to other animals for these studies. In addition, human subjects' ability to articulate makes it possible to conduct studies that address effects on mood. Finally, researchers can use PET scans to study metabolic changes in living subjects without risk to those subjects.(7,24,25)
Another important question is, "What therapeutic interventions can ameliorate AD symptoms?" Knopman has noted, "The absence of a suitable animal model for Alzheimer disease leaves therapeutic trials in human subjects as a necessity."(26) Indeed, many therapies that appeared promising in animal studies have failed to ameliorate AD. For example, choline analogues can reverse deficits in animals, but these drugs have been largely ineffective in people.(9,23,27) Certain cholinesterase inhibitors, such as physostigmine and tetrahydroaminoacridine, may help AD patients.(23,28) Having found a relative excess of acetylcholinesterase and butyrylcholinesterase activity in AD patients' neuritic plaques and neurofibrillary tangles, Mesulam et al. have hypothesized that this excess may account for the mild benefit of physostigmine in alleviating AD memory deficits: "Cholinesterase inhibitors appear to have a major effect directly upon the neuritic plaques and neurofibrillary tangles of patients with AD."(23) Other types of cholinomimetics, however, have not alleviated AD symptoms.(23) The efficacy of cholinesterase inhibitors in treating AD appears to reflect AD's unique pathology. Thus, the benefits of cholinesterase inhibitors such as physostigmine and tetrahydroaminoacridine, as opposed to the disappointing failure of other cholinomimetics, could not be predicted in any investigation not involving AD patients.
Cerebral metabolic enhancers, such as dihydroergotoxine (DHE), have shown promise in double-blind studies. Hutton has noted that DHE may act as an anti-depressant, "as the improvements are most prominent in depression-related symptoms such as emotional withdrawal, depressive mood, and motor retardation."(1) Since other animals cannot articulate their feelings, it is difficult, if not impossible, to study depression by way of animal models, Consequently, the antidepressive effect of different therapeutic modalities requires human clinical investigation. Because depression is a common symptom in AD patients,(1) treatment of depression is important to AD management.
Conclusions
Although widely used, animal models have failed to provide significant insights into AD pathology and management. Kordower and Gash have cautioned, "while there is a desperate need for developing and testing new strategies for AD, the animal models which are currently available mimic only selected aspects of the disease state and leave considerable concerns about their predictive value for designing clinical tests."(13) Similarly, Haroutunian has acknowledged "the discrepancy between the animal studies and the AD pharmacotherapy results."(22)
AD appears to be a uniquely human condition, and all existing animal models differ from AD in important respects. Indeed, our current understanding and treatment of AD is based on clinical data. Human clinical investigation should be the primary focus of AD research efforts.
References
Hutton TJ: Alzheimer's disease: Evolving clinical concepts and management strategies. Comp Ther 1987;13:55-64.
2. Braunwald E, Isselbacher KJ, Petersdorf RG, Wilson JD, Martin JB, Fauci AS: Harrison's Principles of Internal Medicine, 11th Edition. New York; McGraw Hill, 1988.
3. Henderson AS: The risk factors for Alzheimer's disease: A review and a hypothesis. Acta Psychiat Scand 1988;78:257-275.
4. Coyle JT, Singer H, McKinney M, Price D: Neurotransmitter specific alterations in dementing disorders: insights from animal models. J Psychiat Res 1984;18:501-512.
5. Kesner RP: Reevaluation of the contribution of the basal forebrain cholinergic system to memory. Neurobiol Aging 1988;9:609-616.
6. Whitehouse PJ: Specific neurochemical systems and memory. Neurobiol Aging 1988;9:639-640.
7. Henderson VW, Finch CE: The neurobiology of Alzheimer's disease. J Neurosurg 1989;70:335-353.
8. Heston LL, Mastri A, Anderson E, White I: Dementia of the Alzheimer type. Arch Gen Psychiat 1981;38:1085-1090.
9. Smith G: Animal models of cholinergic denervation. Brain Res Rev 1988;13:103-118.
10. Wenk GL: Amnesia and Alzheimer's disease: Which neurotransmitter system is responsible? Neurobiol Aging 1988;9:640-641.
11. Whitehouse JM: The effects of physostigmine on discrimination learning. Psychopharmacol (Berl) 1966;9:183-188.
12. Drachman P: Memory and cognitive function in man: Does the cholinergic system have a specific role? Neurology 1977;27:783-790.
13. Kordower JH, Gash DM: Animal models of age and disease-related cognitive decline: Perspectives on the models and therapeutic strategies. Neurobiol Aging 1988;9:685-689.
14. Walker LC, Kitt CA, Struble RG, Wagster MV, Price DL, Cork LC: The neural basis of memory decline in aged monkeys. Neurobiol Aging 1988;9:657-666.
15. Bartus RT, Fleming D, Johnson HR: Aging in the rhesus monkey: Debilitating effects on short-term memory. J Gerontol 1978;33:858-871.
16. Moss MB, Rosene DL, Peters A: Effects of aging on visual recognition memory in the rhesus monkey. Neurobiol Aging 1988;9:495-502.
17. Lister RG, Sunderland T, Weingartner HS: Psychobiological modeling of age-related memory changes. Neurobiol Aging 1988;9:505-506.
18. Kowall NW, Beal MF, Busciglio J, Duffy LK, Yankner BA: An in vivo model for the neurodegenerative effects of β amyloid and protection by substance P: Proc Natl Acad Sci USA 1991;88:7247-7251.
19. Kotata G: Alzheimer-like effect is reported cut in rats. New York Times Aug 20, 1991.
20. Wirak DO, Bayney R, Ramabbadran RP, et al: Deposits of amyloid β protein in the central nervous system of transgenic mice. Science 1991;253:32.3-325.
21. Bartus RT: The need for common perspectives in the development and use of animal models for age-related cognitive and neurodegenerative disorders. Neurobiol Aging 1988;9:445-451.
22. Haroutunian V, Kanof PD, Tsubayama GK, Carnpell GA, Davis KL: Animal models of Alzheimer's disease: behavior, pharmacology, transplants. Can J Neurol 1986;13S:385-393.
23. Mesulam M, Guela C, Moran MA: Anatomy of cholinesterase inhibition in Alzheimer's disease: Effect of physostygmine and tetrahydroaminoacridine on plaques and tangles. Ann Neurol 1987;22:623-691.
24. Duara R, Grady C, Haxby J, et al: Positron emission tomography in Alzheimer's disease. Neurology 1986;36:879-887.
25. Hoffman JM, Guze BH, Baxter LR, Mazziotta JC, Phelps ME: [18F]-Fluorodeoxyglucose (FDG) and positron emission tomography (PET) in aging and dementia: A decade of studies. Eur Neurol 1989,29(suppl 3):16-24,
26. Knopman D: The necessity of human trials of putative therapies for Aizheimer's disease. Neurobiol Aging 1989;10:633-634.
27. Caine ED: Cholinomimetic treatment fails to improve memory disorders. N Eng J Med 1980;585-586.
28. Summers WK, Majovski LV, Marsh GM, Tachiki K, Kling A: Oral tetrahydroaminoacridine in long-term treatment of senile dementia, Alzheimer type. N Eng J Med 1980;315:1241-1245.