Animal model and evaluation of senile dementia

Section 1 Definition of Alzheimer's Disease

Alzheimer's disease (AD), also known as senile dementia, is a progressive brain degenerative disease or syndrome characterized by aging and characterized by decreased cognitive function. The patient's entire brain is diffusely atrophied with obvious pathological changes in histology—senile plaque (SP) (or neuritic plaque) and neurofibrillary tangle (NFT). Decreased cognitive function in patients with neuronal dysfunction, decreased number of neurons, expanded neuronal volume and synaptic loss (type I dystrophic axons) and formation of intracellular phosphorylated tau (microtubule-associated protein tau) For helical filaments, ie NFT (type II nutritional disorders axons). In addition, extracellular amyloid β-protein (Aβ) is deposited to form SP, which is surrounded by dystrophic processes. The form of Aβ deposition consists of early diffuse aggregation (immature diffuse plaques) to the formation of mature dense plaques consisting of 8 nm diameter fibrils, stained with Congo red and exhibiting birefringence under polarized light. characteristic. Late pathological changes are caused by inflammatory responses involving activated microglia and astrocytes, which surround the deposited plaque. The early behavioral changes of the disease were mild memory disorders or personality changes, gradually worsening within 5 to 10 years, and eventually severe dementia symptoms, life can not take care of themselves. This insidious and destructive brain degenerative disease deprives the victim of the most human-characteristic qualities—memory, reasoning, abstraction, and language.

First, clinical symptoms

Almost all of AD begins to develop in an imperceptible manner, often with the occasional difficulty of recalling recent events in daily life. The patient may not be able to recall a conversation with someone or an activity that has been involved, or may be confused about information about a recently accepted project, which is a precursor to mild cognitive impairment (MCI). . Patients often start with pure symptoms of amnesia, and other cognitive aspects have little or no difficulty [1]. Patients with MCI or early AD remain completely awake, have no significant language confusion, and maintain normal motor and sensory function.

In the first few years of AD, most patients have problems in general cognitive function, such as time and sense of space and complex operations that are correct and free to make mistakes, and difficulties in using words and mathematics. As these minor errors become noticeable, the patient may become boring, indifferent, and emotionally unstable about activities and hobbies. After progressive memory and cognitive decline, many patients began to experience the first, but more obvious, motor dysfunction, including manual surgery, writing, painting, and normal walking. In the course of a few years or a decade, AD will gradually deteriorate to the point of obvious dementia. Due to complete loss of orientation, severe memory impairment, and cognitive dysfunction, many patients are limited in activity, forced to sit or lie down, and may eventually die from certain small respiratory diseases such as aspiration or pneumonia.

Second, neuropathological features

Although the clinical diagnosis of AD patients can be made based on the above-mentioned physical symptoms and symptoms, it is still necessary to perform pathological biopsy after diagnosis. Typical pathological lesions were observed in pathological sections of the hippocampus, amygdala, frontal, temporal, and parietal cortex of AD patients. Although mild atrophy of the cerebral hemisphere (8% to 15%) can be observed by the naked eye, this is not more serious than that of non-demented individuals. However, under light microscopy, the brain regions of AD patients showed SP and NFT. The SP is slightly spherical, and the Aβ fibers are deposited extracellularly, surrounded by degenerated axons and dendrites, activated microglia, and astrocytes. Many of the SPs in which the SP is located can also be seen with the accompanying scattered "plaque" patches. In these small, plaques that are positive for Aβ immunoreactivity, Aβ fibers are usually absent, containing little or only minimal degenerative processes or abnormal glial cells. In most AD patients, the number of diffuse plaques clearly exceeds the number of SPs. The diffuse plaque appears to be the earliest pathological change in the AD brain detectable under light microscopy. They occur before SP, and they also occur in the brains of normal cognitive, middle-aged, and healthy people [2], just as patients with Down syndrome developed into typical AD-like SPs and NFTs. It has the same pathological changes. In fact, there is a morphological continuity between the diffuse plaque in the cortex and the SP, and two types of pathological damage should not be made.

In addition to SP, another indicator for diagnosing pathological damage in AD is NFT. NFT is a double-helical filament structure, usually mixed with straight filaments, located in the cytoplasm around the nucleus of the marginal line and cortical neurons. These abnormal filaments can occur in many SP-destroyed neurons, and some occur outside of SP; NFT can also be in the subcortical nucleus, such as the medial septal nucleus and the basal forebrain Meynert nucleus (NBM) cholinergic nerve. Observed in Yuanzhong. These nuclei emit fibers that are widely projected into the edgeline and joint cortex rich in Aβ deposits.

A recent study of AD found that SP is also seen in individuals with no or only mild cognitive impairment. However, it is important that the type of SP in almost all normal aged brain tissue is diffuse, that is, these plaques lack the neuronal and glial pathological changes associated with Aβ deposition, while there is no or very Less NFT. Based on these findings, it is speculated that the plaque is a "preclinical" lesion, and neuronal damage and its processes cannot be observed under microscopic observation, just like the "fat lines" appearing on the wall of the majority of asymptomatic elderly patients. It is the same as the precursor of clinically important, mature atherosclerotic plaque.

Third, neurotransmitter changes

The first neurotransmitter abnormality found in AD brain tissue is acetylcholine (Ach), and the enzymatic activity of synthesis and degradation of Ach is altered [3]. Therefore, the number and quality of the medial septal nucleus and basal forebrain cholinergic neurons of AD vary. However, the lack of these transmitters is accompanied by the reduction or loss of other neurotransmitters, including glutamate (Glu), gamma-aminobutyric acid (GABA), somatostatin (SST), and adrenal gland. Corticosteroid releasing factor (CRF) and serotonin (5-HT) and so on [4]. Therefore, AD neurotransmitter deficiency is extensive and there are no clear clues about the disruption of various neuronal subtypes. Until now, only cholinergic transmitters have been used effectively for AD treatment.

Section 2 Alzheimer's Animal Model

Because the etiology and pathogenesis of AD is very complex, it is the result of the interaction of environmental factors and genetic factors. It is a multi-cause disease or clinical syndrome characterized by progressive cognitive impairment. It has its characteristic neuropathological table. Type and neurotransmitter phenotype. Therefore, there is currently no ideal animal model fully equipped with AD characteristics, which greatly restricts the research of AD therapeutic drugs. So what kind of AD model is the ideal AD model? According to the above characteristics of AD, the author believes that the ideal animal model of AD should have the following three characteristics: (1) the main neuropathological features of AD - SP and NFT; (2) the occurrence of brain neuron death, synapse Significant pathological changes in AD such as loss and reactive gliosis; (3) cognitive and memory dysfunction. If any model can meet the above characteristics at the same time, it will be a good AD model. There are many types of animal models, but most of them only simulate a part of AD. It is difficult to fully possess the characteristics of an ideal animal model, and it is difficult to fully simulate the characteristics of AD disease. The transgenic animal model that has emerged in recent years is a hot spot, but it cannot completely replicate the characteristics of AD. The existing model is described below.

First, the aging-based AD animal model

In the history of biomedicine, the topic of AD has hardly caused much interest among scientists. This situation changed dramatically in the second half of the 20th century, when the per capita life expectancy in the United States suddenly increased from 49 to 76, increasing the age of individuals to the age at which neurodegenerative diseases prevail. AD occurs in the elderly, and aging is a positive risk factor for AD. With the increase of age, the prevalence of AD increases exponentially. It is generally believed that over 65 years of age, the proportion of patients with AD who are 5 years old increases doubles. At present, the world's population is aging, and the incidence of the disease is further increased, making it an increase in the proportion of the elderly. AD has become the most common cause of dementia over the age of 65 and has been recognized as a major public health problem in developed countries.

The living body is different from the general objects in nature. It can metabolize. It can replace the old chemicals in the body with new chemicals. Why does aging happen and how does it happen? At present, it is believed that aging may be related to the damage of oxygen free radicals to cells. Some people think that it may be caused by energy metabolism disorder caused by mitochondrial dysfunction. It is also thought to be related to the role of a certain gene controlling senescence. If human beings can understand the mystery of aging, then they will understand the mechanism that aging makes AD more likely to occur. Maybe people will have AD at a certain age, and those who do not have AD may not live yet. To his or her age of AD. This type of model is based on aging as the basis of AD, and through various methods to promote animal aging (including natural aging) to achieve the purpose of making animal models of AD.

(1) Animal models of natural aging cognitive impairment

Animal models of AD obtained by natural aging of animals themselves, including aged rats, mice, and monkeys. Neurological changes such as cognitive impairment in such models occur naturally, and are closer to the true pathophysiological changes of AD than other animal models [5]. However, animals are difficult to obtain in large quantities, and older animals are prone to death, and there is no guarantee that each has the characteristics of AD.

(2) senescence accelerated mouse (SAM) model

Juntian Zhutian [6] obtained a natural fast-aging mouse by inbreeding AKR/J naturally-mutated mice. SAM P/8 and SAM P/10 in many strains of this family showed obvious learning and memory. The function is diminished, in a state of dementia with low tension and low horror.

(3) D-galactose-induced subacute aging model

The D-galactose damage model was first proposed by Chinese scholars. Animals showed signs of aging such as learning and memory loss, slow movement, and sparse hair. The organelles in cortical neurons are reduced, mitochondrial expansion is vacuolar-like degeneration, rough endoplasmic reticulum degranulation, protein synthesis is reduced, and neurons are lost, which is consistent with the performance of older animals [7].

The animal model obtained by simulating the aging process more realistically reproduces the pathophysiological changes of AD, which has a certain similarity with AD. However, aging is only a risk factor for the onset of AD. It does not guarantee that the aged animals will develop into AD, and generally do not have the characteristics of SP and NFT of AD, and are not easy to raise and survive.

Second, the animal model of AD based on the theory of cholinergic

In the mid-1970s, the first neurochemical cues about AD came from the severe degeneration of neuronal synthesis and release of Ach, which was found to be involved in Ach synthesis and degradation in the cerebral cortex. The content and activity of transferase (ChAT) and acetylcholinesterase (AchE) are decreased, and these regions are accompanied by deletion of cholinergic neuronal bodies. In other words, the medial septal nucleus and the base forebrain cholinergic system function are reduced. This finding has focused pharmacologist research on how to increase the level of Ach in the synaptic cleft, primarily by inhibiting the degradation of Ach. The end result of these efforts led to the use of cholinesterase inhibitors for the treatment of AD.

It has been confirmed that the activity of the cholinergic system is closely related to the process of learning and memory of human beings. Basal forebrain cholinergic neurons, hippocampus and cortex and pathways between them are important structural basis for learning and memory function. Significant injury or death of basal forebrain cholinergic neurons, synthesis of presynaptic Ach, activity of ChAT, and uptake of choline in AD patients were significantly associated with the degree of cognitive impairment in patients with AD. [8]. Therefore, it is considered that the degradation of the cholinergic nervous system in the brain is one of the main reasons for the decline of learning and memory function in AD. In summary, cholinergic dysfunction in AD patients is closely related to learning and memory, and is one of the important reasons for dementia. Therefore, the cholinergic theory of AD has gradually formed. Based on this theory, people use various methods to destroy the function of the cholinergic system of the animal's brain, and to promote learning and memory impairment to achieve the purpose of making animal models of AD.

In mammalian brain, basal forebrain NBM is the main site of cholinergic neurons, and 70% of cholinergic projection fibers in cerebral cortex and hippocampus are from NBM. According to the theory of cholinergic injury in the pathogenesis of AD, the occurrence of AD is caused by a decrease in the function of the medial septal nucleus and the basal forebrain cholinergic system. This model destroys this region by local injection to prepare basal forebrain cholinergic energy. AD model of system damage. This model mimicked cognitive deficits in AD and extensive functional impairment of the basal forebrain cholinergic system. Such models are:

(1) Electrical damage, surgical damage

By performing surgery on the animal, referring to the stereotactic map of the animal brain, the NBM is damaged by electric burn, and the hippocampus dome is surgically cut off. After surgery, the animals had learning and memory dysfunction, but there was no SP and NFT in the pathology, and this method has a large damage range, which is currently not used [9,10].

(2) Chemical damage

Excitatory amino acids such as ibotenic acid (IBO), kainic acid (KA), quisqualic acid (QA) and N-methyl-D-aspartate (***A) The basal large cell nucleus injected into the animal can establish an AD model [11]. AF64A is a cholinergic nerve ending-specific neurotoxin with a structure similar to that of choline. It is theoretically only taken up by cholinergic neurons and selectively acts on the high affinity choline transport (HAChT) system. Stereotactic injection is used to damage the perinuclear body of neurons without destroying the nerve fibers passing through this area, so the selectivity of damage is better, the animals have memory dysfunction, but there is no pathological change of SP and NFT [12] , 13].

From the late 1970s to the early 1980s, as other neurotransmitter deletions were discovered and identified in AD brain tissue, the pathogenesis of AD became more and more clear. It was different from Parkinson's disease, not a single one. The neurotransmitter degenerative disease is caused by the degradation of multiple transmitter systems that are mixed together. People's attention has gradually focused on the mechanisms of synaptic dysfunction and perinuclear deterioration, as they affect many types of neurons in the limbic and combined cortex. This understanding may explain why most patients taking cholinesterase inhibitors have a poor long-term clinical outcome. For the above reasons, this model only replicates the characteristics of AD cholinergic dysfunction, does not replicate the characterization of other transmitter dysfunction, and does not show the typical pathological changes of AD - SP and NFT.

Third, the transgenic animal model based on the genetics of AD

(a) Aβ hypothesis

Since Alzheimer first reported AD patients with progressive memory and cognitive decline in 1906, the disease named after the discoverer has been plaguing people. Alzheimer noted that the patient's brain had NFT and extravascular SP pathological changes, but could not distinguish whether it was a causative agent or merely a marker of disease. It was not until the end of the 20th century that with the advancement of molecular biology methods and related sciences, people became more aware of AD. In 1991, an important clue was found in gene-related analysis: missense mutations in the amyloid beta-protein precursor (APP) gene lead to autosomal dominant familial AD (FAD) Occurrence, these mutations occur in the Aβ coding region of APP [14]. Subsequently, three other genes related to the pathogenesis of AD were found, the presenilin-1 (PS1) gene and the presenilin-2 (PS2) gene on chromosomes 14, 1 and 19, respectively. The apolipoprotein E4 (apoE4) gene [15, 16, 17], their mutation or polymorphism increases the risk of developing AD. The APP, PS1 and PS2 genes are mainly related to FAD. FAD patients have at least one of them abnormal. About 10% of FAD and 2% of sporadic AD have APP gene mutations, while those with PS gene mutations are as high as 40% to 50%. [17]. ApoE is associated with both FAD and sporadic AD. The clarification of every genetic change associated with FAD and the change from genotype to phenotype has led to the consensus that at least the formation of FAD begins. Therefore, people turned their attention to the formation of SP, the main component of Aβ. After a series of studies, some of the metabolic mechanisms of key enzymes α, ß, γ secretase (secretase) and their molecular mechanisms were obtained. progress. These new findings, combined with the prevalence of Aβ, are neurotoxic proteins that cause neuronal dysfunction in the brain and the corresponding clinical manifestations of AD. A rough outline of the disease development process began to emerge, gradually forming the Aβ hypothesis of AD. The theory suggests that the end result of various causes is the abnormal metabolism of Aβ, which is characterized by excessive production of Aβ, reduced degradation, and accumulation or deposition of Aβ. , triggering a variety of immune inflammatory reactions and neurotoxic cascades, leading to a wide range of neuronal degeneration and cell dysfunction, synaptic loss and neuropathy caused by apoptosis or death, a series of pathophysiological changes, ultimately leading to AD . Therefore, Aβ deposition is the initiating factor and central link of AD pathogenesis [18], which naturally becomes the main direction and hotspot of AD research, and also brings new drug targets for the treatment of AD.

(2) Transgenic and AD transgenic animal models

At the same time, it was found that the mutations in the above AD genes began to use the newly emerged transgenic technology to transfer the exogenous genes of the four human mutations that have been discovered into the animal to prepare a transgenic animal model, which has become a model in animal model making. A relatively new method for transferring FAD-related human mutant genes into animals, and allowing the exogenous genes to be stably inherited, altering animal genetic traits, and achieving the purpose of mimicking the genetic characteristics of human FAD in animals. Screening for AD treatments for A[beta] targets.

A transgene is a method of introducing an exogenous gene into a genome and stably inherited the exogenous gene. In the past 15 years, the transfer of exogenous genes into the mammalian genome has become a routine experimental method. The traditional method is to introduce the target gene into a single cell fertilized egg by microinjection, and the surviving fertilized egg is transplanted into the uterus of the pseudopregnant animal to develop into an individual. When part of the fertilized egg has not yet split, the exogenous DNA has integrated and entered the animal genome. An animal developed from such an egg can stably inherit an exogenous gene to a progeny because its exogenous gene is integrated into the germ cell. With the advent of embryonic stem (ES) technology and the development of homologous recombination gene targeting strategies, researchers can explore the function of specific genes and understand the effects of fine-grained changes on gene function and production in vivo. Specific effects. This new effective transgenic technique is to first establish an ES cell line, remove cells from the developing blastocyst inner cell mass, and upload on the trophoblast or medium with differentiation inhibitory activity to maintain the dedifferentiation state, and then by electric shock. , transfection, microinjection and other methods to transfer exogenous genes into ES cells, cells are screened and then sent to blastocysts, blastocysts transplanted into pseudo-pregnant females to develop into individuals. Individuals born in this way are mostly chimeras compared to individuals who have been microinjected with traditional single-cell fertilized eggs, since the cells carrying the transferred gene account for only a portion of the entire blastocyst. As long as cells carrying exogenous genes can develop into germline cells, transgenic lines can be established by breeding. Introduction of exogenous genes into ES cells is superior to microinjection of single-cell fertilized eggs because foreign DNA can be designed to homologously recombine with endogenous homologous sequences. After Southern blot or PCR screening, the correctly targeted ES cell clones can be identified and developed into the desired chimera after blastocyst injection into the host blastocyst.

Transgenic animals can study the role of a specific pathogenic gene in vivo and are a unique and important model for studying AD. Transgenic animals fall into two categories, the first being mice that contain a null gene, the knockout mouse [19,.20]. They are commonly used to understand the function of knockout genes in the etiology of AD, while they can re-insert relevant human mutant genes to avoid possible interference with endogenous mouse genes.

The second category is the disease-causing gene of AD transgenic mice expressing AD mutations, thereby showing the pathological features of the associated AD phenotype, by adding additional genes to the existing AD transgenic mice, or creating a double Gene or multi-gene mice. This allows people to study the interaction of these different AD-related factors [21, 22]. Based on genetic background, reproductive capacity, operational difficulty and economic considerations, mice are currently used in transgenic animals. A variety of transgenic mice can express genes associated with AD lesions such as APP, C-terminal fragment of APP, tau, PS1, PS2, ApoE and the like.

1. Parameters required for preparation of Aβ-deposited transgenic mice

When constructing a transgenic mouse, the choice of promoter plays an important role in determining the expression pattern of time and space and the copy number of the gene. There are 15 different promoters for mouse trans-APP and PS1 gene expression. These mice have successfully replicated some of the pathological features of AD, such as overproduction of Aβ and amyloid deposition. Some are also produced by transgenic promoters and are commonly used, such as APP promoter [23], prion protein promoter [24] and PDGF-β [25], which lead to some expression outside the nervous system, while Thy-1 [ 26] The promoter is neuron-specific. The choice of promoter also affects the successful replication of Aβ deposits to establish a parameter for transgenic mice—overexpression of the transferred gene. Early transgenic mice used the APP allele of FAD but did not show any pathological features of AD. This is due to the low level of transfer into the APP gene, probably only 1.3 to 1.4 times that of the endogenous mouse APP gene [27].

The second is the selection of the mouse line, as at least 8 different murine lines were used to make transgenic mice overexpressing APP [24, 26] and expressing PS genes (PS1 and PS2) [28]. Except that the APP and PS1 transgenic mouse lines were generated from the FVB/N strain [15, 47, 72], all other mice transfected with APP and PS1 were established on the hybrid mouse line. Hybrid mice can produce larger volumes and have more robust embryos [29]. The genetic background of the murine used in the transgene will have a significant impact on the maintenance and behavioral test results of the murine line. Because of the undesirable characteristics of hybrid murine lines, transgenic animals will carry these characteristics when they are used. These adverse features include impaired visual acuity in FVB/N, CBA and C3H genetic backgrounds, high tumorigenicity in DBA and 129 murine backgrounds, a higher level of aggressiveness in SJL hybrid mouse lines, and age-dependent deafness in C57 mice [30] ]. In addition, C57 mice showed age-dependent pathological damage aggregation, similar to amyloid deposits, but could be distinguished from them [31]. Hybrid mice with these lesion characteristics can affect their performance in learning and memory tests that rely on the visual system, such as the Morris water maze. For behavioral testing, including learning and memory, the best mouse selection is the C57B1/6J hybrid mouse line. This mouse system can better adapt to the Morris water maze, the eight-arm maze and the conditional space transformation test [29]. Carlson et al [32] studied the effects of the genetic background of mice on their phenotype, and found that Tg2576 mice with SP6xSJL hybrid genetic background can produce SP transgenic to each C57B1/6J mouse. The survival rate of both the descendants and the descendants has decreased. The C57B1/6J background reduced the viability of transgenic mice, and their ability was between APPswe mice and C57B1/6J mouse lines cultivated by Borchelt et al. The data indicated that the genetic background was related to survival status, behavior and overexpression of APP. Pathology has a significant impact.

Before the establishment of a transgenic mouse model that successfully replicated Aβ deposits, it was predicted that there were some problems: ectopic expression of the transferred gene, lack of complete APP isomers, and lack of downregulation of APPs-α. To address these potential problems, yeast artificial chromosome (YAC) was used to generate transgenic mouse lines. Lamb et al [33] established transgenic mice by introducing 400-kbp DNA encoding human APP gene, which expressed all APP splicing subtypes in the correct expression pattern; the second problem is that it may be in mice. It only replicates the hydrolysis of normal APP and the reduction in APPs-α production. APPs-α has been reported to have neurotrophic effects [35, 36] and neuroprotective effects [37]. To test this hypothesis, the sequence containing the Swedish mutant Aβ was substituted for the Aβ sequence in the transgenic mouse line [38]. The advantage of this method is that it does not increase the expression of APP, its expression does not exceed normal levels and no exogenous APP expression is produced. Compared with normal human aging brains, A? production in the brain of this transgenic mouse at 1.3 to 4.5 months of age increased nine-fold. When the APPswe mutation is present, it is estimated that the APPs-α production of these mice may be reduced. It has been observed in these mice that increased production of A[beta] and decreased secretion of APPs-[alpha] have a potential role in neuronal survival. Further evidence suggests that reduced APPs-alpha secretion may play a role in successful replication of the AD mouse model. A recent study showed that intraventricular injection of APPs-α in Swiss mice improved memory capacity [39]. Therefore, potential neuropathological effects of increased Aβ and decreased APPs-α should be carefully monitored in future transgenic mouse studies.

2. APP and PS transgenic mice

(1) Aβ deposition

Early onset AD accounts for a small proportion of all AD patients, mainly occurring between 30 and 60 years of age, usually with familiality. Mutations in three genes have been found in FAD to cause AD, the APP, PS1 and PS2 genes. Transgenic mice overexpressing APP, PS1 and PS2 alleles of FAD can increase brain Aβ levels. Four transgenic mouse lines expressing the FAD mutant APP allele have been established to produce SP with birefringence and age-dependent characteristics [25, 26, 30, 40]. The first report to increase Aβ levels in mouse brains was derived from PDAPP mice transfected with human APP695swe [24] and APP717V-F mutations. The APP level is 2 to 3 times that of endogenous APP in mice, and there is formation of Aβ deposition [25]. YAC technology and human APPswe (R1.40) mice expressing three times the level of murine APP also showed an increase in total Aβ production and a proportional increase in Aβ42 (approximately 20% of the total) [41], compared to previous The results of in vitro transfection of the APPswe cell line were similar [42]. Further analysis of PDAPP mice [25] showed that the regional levels of APP and APPs-β were constant at all ages, while Aβ levels were higher in some regions of the brain than in other regions, and deposition of Aβ in these regions accumulated with age. Consistent with the expected results, Aβ42 produced by APPV717I mutant PDPP mice is the major class of Aβ [43], and similar results have been reported in the cortex of AD patients with APPV717I mutation [44]. There are also such transgenic animal models in China, and over-expressing APP695 and 751 mice produced by Qinchuan [45]. Compared with the control group, Aβ42 immunohistochemistry showed Aβ deposition in the neurons of cerebral cortex, mouse and hippocampus. Congo red staining can form amyloid formation between the cerebral cortex and the cortex.
A comparative analysis of APP levels in several different brain regions of heterozygous and homozygous PDAPP mice indicated that the level of full-length APP in the thalamus of homozygous mice was higher than in the hippocampus of heterozygous mice. Even if the transferred APP gene is continuously overexpressed, the level of Aβ produced by the unit full length APP in the cortex and hippocampus is the highest. The deposition of Aβ occurs in these regions but does not occur in the thalamus [43], suggesting that regional factors in the brain of transgenic mice make it easier for APP to metabolize Aβ and amyloid plaques in these regions. Similar to AD patients, the pathological damage of the cerebellum is the lightest until the late stage of the disease. A notable exception is the PS1E246A mutant mouse whose cerebellar pathological changes can be observed at a relatively early stage of the disease [46].

The level of Aβ42 in the brain of mice overexpressing the PS1M146L or M146V gene was approximately 30% higher than that of Aβ42 expressing PS1 wild type (PS1WT) mice, while there was no significant difference in Aβ40 levels. These data are consistent with the results of increased levels of A[beta]42 observed in FAD containing PS1 mutations. The level of Aβ42 in the brain of mice transfected with the PS2 mutant gene was also higher than that of mice transfected with PS2 wild type (PS2WT). However, even in the study using the same sandwich ELISA assay, in the Oyama et al study, the Aβ level of the transgenic PS2 mice was not consistent with the PS1 transgenic mice described by Duff et al. [47]. Another inconsistency is that the level of Aβ in non-transgenic mice is higher than in PS2 transgenic mice [48], while the overexpressed mutant PS mice do not show any Aβ deposition in the brain, at least in 12 When the age of the month. No report on Aβ deposition occurred in transgenic PS1 and PS2 mutant mice, which may be attributed to the fact that Aβ or mouse Aβ produced in these mice differs from human Aβ in the third amino acid residue. It has been reported that rodent Aβ is not amyloid in humans like human Aβ in vitro, and excessive production of murine Aβ in transgenic mice has been shown to co-immunize with human Aβ as a diffuse deposition [49].

To investigate the effect of the mutant PS1 gene on human APP metabolism in transgenic mice, several research groups have established double transgenic mice. Mouse mice expressing both the FAD PS1 and APP genes showed accelerated amyloid deposition [50]. The progeny produced by transfection of APPswe mice [24] and PS1M146L mice [47] produced sulphur-S-positive Aβ deposits in the cortex at 13 to 16 weeks [43]. Similarly, transgenic APPswe mice and PS1A246E mice hybridized to produce accelerated Aβ deposition, which can be detected at 9 months of age [51]. Mutations in PS1FAD can increase the production of Aβ42 [52]. The accelerated deposition of diffuse plaques in these mice is mainly composed of Aβ42, which is consistent with the hypothesis that these plaques are precursors to AD and Down's syndrome SP. In addition, it was observed that the transfected Thy1-APPswe mouse, APP23, produced Congo red-refractive plaques in the presence of diffuse plaques, suggesting that maturity from diffuse to dense plaques in mice may not be a prerequisite for dense plaque formation. Condition [26]. These in vivo experiments confirmed that mutant PS1 can affect the metabolic process of APP, consistent with in vitro transgenic cell assays. These studies also showed that transmutation of PS and APP hybrid mice provided the fastest pathway for the formation of Aβ deposits in the brain of mice.

To investigate the effect of reduced PS1WT on mouse A[beta]42/43 levels, PS1 knockout mice were mated with trans APPswe mice. There was no significant difference in Aβ levels in the brains of the offspring between January and May compared with PS1WT mice. This suggests that the mechanism by which FAD-associated PS1 mutations cause disease cannot be attributed to a decrease in PS1WT during aging [53]. In another independent study, it was found that the levels of Aβ40 and Aβ42 in PS1 null mice were decreased, indicating that the increase in Aβ42 in AD patients was not caused by the loss of PS1 function due to mutations in PS1. Mice that introduced the Thy1-PS1A246E gene in the null background of the PS1 knockout also measured Aβ, and the level of Aβ42/Aβ43 was increased compared to the Thy1-PS1 WT transgenic or non-transgenic control [21].

(2) Change of Tau

The existing nerve fibers of the APP and PS gene mice were analyzed by silver staining, and no NFT was detected in the FAD allele mice overexpressing PS1 and APP. However, in some murine lines there was an early change in NFT - hyperphosphorylation of tau. In Thy1-APPswe mice, several phosphorylated epitopes were confirmed using the tau antibody and it was found that the distorted axons contained highly phosphorylated tau [26]. NSE-APP751 mice have also been reported to stain positive for Alz50 [54].

(3) Atrophic axons, neuronal loss and reactive gliosis

In PrP-APPswe mice [24], Gallyas silver staining was used to find atrophic axons in the vicinity of the plaque. Distorted axons observed in anti-synaptophysin antibodies in PDAPP mice are similar to axonal changes in AD patients, with dystrophic axons with dense plate-like and neurofilament aggregation, with AD dystrophic dystrophies Axon features [25]. Sturchler-Pierrat [26] also observed a distorted axon around the core of Aβ deposition using a specific method and found that there was a distorted AchE-positive fiber locally, and cholinergic neurons in AD patients are currently known to be severely damaged. of.

Synaptic loss and neuronal death are typical features of AD. Loss of neurons can occur in the vicinity of SP, which has been reported in PDAPP717 [25] and Thy1-APPswe [26] mice. However, stereometric studies of 18-month-old PDAPP mice showed that plaques in the cortex, hippocampus or buckled band did not show significant neuronal loss; synaptophysin, MAP-2, cytochrome oxidase The levels of -2 and cytochrome oxidase-4 were not reduced, while the level of glial fibrillary acidic protein (GFAP) was significantly increased [55]. Similarly, there was no significant reduction in the number of neurons in the cortex and hippocampus of PS1A246E transgenic mice in September and December [51].用相同的研究方法发现14到18月龄的APP23（Thy1-APPswe）转基因小鼠，其CA1区的金字塔神经元有明显的减少 （14-25%），与此区域斑的数量呈负相关。其新皮质区域神经元计数与斑的数量之间却没有相关性。将同样的技术应用于AD病人大脑中得到了一个相似的结 论[56]。 APP23小鼠和其它APP转基因小鼠之间的区别是APP23有一个更高的致密斑数量（占到90%）。至今没有报道表明PS1和APP双转基因小鼠中出现有神经元损失的病理改变。
反应性胶质增生已在一些小鼠模型中报道。在PrP-APP695swe小鼠 [24]、PDAPP小鼠[55]和YAC过度表达APPsweR1.40系转基因小鼠（Lamb et al.NSc.1998）中报道过GFAP染色阳性的星形胶质细胞增生，用细胞形态学方法检测到小胶质细胞增生以及轴突的异常改变。但仅是在那些含有很高 刚果红折光性斑的小鼠中才检测到炎性反应[26]。反应性胶质细胞增生并没有在9月或12月龄的HuPS1A246E小鼠检测到，推测是由于Aβ沉积缺乏 所致。在PS1-A246E/APPswe和PS1 M146L/APPswe双转基因小鼠中，使用抗GFAP的抗体能检测到反应性星形胶质细胞包围着Aβ沉积[51]。因此，反应性胶质增生似乎与Aβ沉积 相联。

（4）记忆力障碍

在小鼠的行为学研究中，Morris 水迷宫经常被用来检测空间参考记忆；Y迷宫去检测空间近期记忆；而放射状迷宫用来检测空间工作记忆。采用这些测试方法，在APP转基因小鼠上发现学习和记 忆障碍。Tg2576小鼠在10月龄时，与对照的非转基因小鼠相比，在Y迷宫和Morris水迷宫显示出明显成绩下降。Tg2576小鼠回交到SJL背景 小鼠的后代12-15月龄时在Morris水迷宫测试中在出现记忆力障碍。这些小鼠的行为和其双亲Tg2576鼠相似[24] 。这表明年龄依赖的在行为学的改变不受小鼠遗传背景影响。

从Tg2576与APPswe杂交获得的双转基因小鼠[57]在3-3.5月龄时，在Y迷宫测试中成绩明显下降。这个双转基因的小鼠中发生的行为学障碍并 不比Tg2576 和APPswe小鼠发生得更早。Tg2576双亲鼠系3月龄的更早期分析表明，Y形迷宫测试中几乎有确定的数值改变，但数据却没有统计学差异，这可能是由 于研究的老鼠数量偏少的原因[24]。这表明Tg2576小鼠在Y迷宫的出现的行为学改变既不是年龄依赖性的，也不因PS1M146L基因突变的存在而加 速。NSE-APP751小鼠同样在水迷宫测试中表现出空间学习障碍。其行为学障碍在6-12月龄时变得明显起来[58]。这表明认知的损害是由于APP 的过度表达引起的，而此时还没有淀粉样沉积的出现。与此相似，PrP-APPswe鼠回交到C57B6/L背景中制造出的小鼠用Morris水迷宫检测， 其行为学障碍发生于12月龄，与对照组相比具有更长的逃避潜伏期[59]，而这些行为学障碍的小鼠却缺乏Aβ沉积。这些小鼠产生最早的Aβ沉积部位是在齿 状回分子层外部，与AD病人大脑SP经常发生在齿状回分子层外部相似。综上所述，转基因小鼠在Morris水迷宫和Y迷宫所表现出认知功能的下降开始于任 何分子层侦测到淀粉样沉积以前。需要进一步确定的是，淀粉样沉积是否进一步加重了小鼠在Y-迷宫中的行为学损害。

3. Aβ和C100-4转基因小鼠

转基因小鼠也通过使用Aβ，100或104个氨基酸的APP片段来产生。C100是β分泌酶剪切APP的产物，能够被γ分泌酶所水解而产生C端的Aβ。 FVB/N转基因小鼠过度产生Aβ42，表现在整个皮质、齿状回、丘脑和后脑的神经元都有Aβ的免疫反应[60]。和Hsiao APPswe FVB/N小鼠相似[52]，这些FVB/N小鼠表现出神经系统的异常和幼年的死亡，用TUNEL染色方法检测到神经细胞凋亡和神经细胞发生的变性 [60]。推测可能是过多的Aβ42直接引起了神经元的变性和凋亡，导致这些小鼠身上所观察到的神经学上的异常，但也可能是FVB/N遗传背景产生的特殊 表现[32]。

三个研究小组产生了过度表达APP片段C100（或C104）的转基因小鼠[61，62，63]。APP的C100区域含有APP的跨膜区域和α、β、γ 分泌酶的酶切位点。有一个是与信号肽设计在一起，以确保C100片段进入细胞的膜内室，在那里与分泌酶发生作用。C100的表达在启动子的控制下制造出年 龄依赖的神经变性和海马齿状回区域的突触损失。表达C104的B6C3鼠使用的是神经丝轻链启动子，大脑结构中检测到年龄依赖的Aβ聚集的免疫反应。然 而，在免疫杂交和免疫沉淀中却没有检测到4kD的Aβ蛋白。这些小鼠海马CA1区表现出年龄依赖星形胶质、小胶质增生反应和神经元损失。其在Morris 水迷宫中出现了空间学习障碍以及减少的长时程增强（LTP），而长时程抑制（LTD）却没有变化[62]。以上结果中产生出这样的问题，究竟APP哪一方 面的代谢产生神经变性和认知障碍？一些转基因小鼠发展成行为学障碍而没有Aβ沉积的形成，这表明神经元功能障碍起源于APP的水解及运输与Aβ的产量有关 联，但却不依赖于Aβ的产量。相反，在含有一个信号肽转基因的C100小鼠，Aβ的产量能在C57BL/6×DBA鼠中检测到[63]。该鼠到9月龄时没 有产生明显的淀粉样沉积，也没AchE活性的下降。在这些小鼠中，Aβ的沉积是否呈现年龄依赖的模式将是一个值得深入研究的问题。

4.载脂蛋白E小鼠

ApoE4是与AD相关联的危险因素。ApoE4对含有APP突变的FAD有促进作用，但对具有PS突变的FAD却没有作用[64]。ApoE4与AD的 发病相联不仅在病例对照研究所支持[65]，而且家族性AD为基础相关研究中得到支持[66]。.现在认为，晚期发病的AD形式（发生于65岁以后）更为 普遍，这类AD病人的发病主要是与ApoE4有关，认为其参与了β淀粉样斑的形成。含有ApoE4表型的AD病人大脑中Aβ40和Aβ42免疫反应斑的水 平都比不含有ApoE4表型的AD病人的水平要高[32] 。

缺乏ApoE的小鼠与PDAPP小鼠杂交的后代表现出Aβ沉积显著的减少[67]。纯合子的PDAPP+/+ApoE+/+的小鼠6月龄时，海马与新皮质 出现了许多Aβ沉积，而PDAPP-/-ApoE-/-小鼠仅出现稀少的Aβ免疫反应。与PDAPP+/+和ApoE+/+鼠相比，PDAPP+/+ ApoE-/-小鼠的发生的沉积都是弥散斑。这个结果不能归因于改变的APP蛋白水平和APP代谢产生的Aβ，因为在2月龄时，两组的APP与Aβ水平没 有差别[67]。虽然它的分子机制还不是很清楚，却在淀粉样假说和ApoE4危险因子之间建立了联系。

ApoE敲除小鼠的记忆障碍和胆碱能神经元异常表明了ApoE和AD之间的另外一种联系。携带有ApoE4等位基因的AD个体中，观察到AchE神经元的 减少[37]。而缺乏ApoE的小鼠，通过生物化学和免疫组织学测量发现其海马与皮质中的AchE活性下降，其在Morris水迷宫的工作记忆下降 [68]。这种在Morris水迷宫中的空间工作记忆障碍和胆碱能标记物的下降能够被M1激动剂治疗3周所逆转[69]。然而，第二个ApoE敲除小鼠胆 碱能神经元研究发现，与配对的C57B1/6野生型小鼠相比，没有明显的胆碱能神经元数量和大小的减少及AchE活性的下降[70]。前一个研究中报道的 胆碱能缺陷可能是选择小鼠的遗传背景影响了胆碱能神经元生物特性 [71]。如果鼠系的精确遗传背景与已知和合适对照小鼠相比，那么从一个遗传背景到另一个遗传背景的结果不相一致性，最终是能提供有用的信息，这需要不同 遗传背景中确立一个遗传背景作为标准通用的转基因鼠系。人类遗传疾病中个体不同外显率可能是由于遗传背景中突变的相互作用造成的，可能对于某些危险因子， 如ApoE4基因的作用特别明显。ApoE-/-小鼠中的载脂蛋白基因敲除并替代成ApoE的异构体——ApoE3或ApoE4已表现出其在水迷宫测试中 出现的记忆力损害。在表达ApoE4的6月龄小鼠比对照小鼠或表达ApoE3小鼠，在水迷宫测试中在寻找隐藏平台方面有一个更长的潜伏期[72]。

5. Tau基因小鼠

AD两大病理改变除了SP外，另一个是NFT，其由高度磷酸化的微管相关蛋白tau构成的。正常的情况下，tau蛋白分布于轴突，结合微管和聚合的微管蛋 白，在轴突微管之间建立短的连接桥。AD病理条件下，tau被磷酸化重新分布于胞体和胞体树突室，在那里聚集形成丝状物（PHFs，pair helical filaments）。在疾病的晚期阶段，这些丝充满整个神经元胞体（NFT，neurofibrillary tangles）。SP和NFT能各自独立的发生，由tau蛋白聚集形成的纤维缠结在一些情况下与那些普通神经变性疾病中的纤维缠结难以区别，但在这些普 通神经变性疾病中几乎没有发现含有Aβ沉积物和SP。相反的，Aβ的沉积物能在认知功能正常的老年人中发现，但却没有纤维缠结的发生。也有一些不常见的 AD病理是“纤维缠结稀少”的类型，即很少的NFT而有丰富的Aβ沉积斑。由磷酸化tau构成的NFT存在于一些不含SP的神经退行性疾病，如 Pick's 病、渐进性核上性麻痹、皮质基底退行性疾病和17号染色体相联帕金森综合征额颞叶痴呆（frototemporal dementia with Parkinsonism linked to chromosome 17，FTDP-17）。这些疾病的NFT分布不象AD只发生于神元内，还分布于胶质细胞中。到目前为止，在AD病人中并未发现在tau基因存在的异常改 变，而最近在FTDP-17中却发现了具有致病的tau基因突变[73]。

（1）人类野生型tau基因小鼠

目前APP和PS转基因模型小鼠不能形成tau丝，而当时与FTDP-17相联的tau突变还未发现，几个研究小组制作了人类野生型tau转基因小鼠。第 一个tau小鼠模型[74]表达的是长的四倍（four-repeat，4R）tau同型体——htau40，采用的是人Thy1启动子。使用全脑免疫杂 交的方法，发现人的tau基因的表达水平为内源性鼠Tau的10%。带有人的tau标记的神经元能在大脑大部分区域观察到，但它们的数量仍相对的少，只占 整个神经元的一小部分。和AD的一样，转基因鼠中人的tau存在于神经元胞体、轴突和树突。其磷酸化位点包括Ser202/Thr205和Ser396 /Ser404，能被针对相应磷酸化位点的tau蛋白抗体AT8和PHF1所确认。转基因鼠没有观察到tau蛋白丝的形成，而染色的tau为均质或粒状 的，而不是纤维状。另一个使用的是鼠3-羟基-甲基-戊二酰CoA还原酶启动子的转人最短tau同型体——htau44的小鼠也有相似的结果。总之，这些 人类野生型tau基因小鼠显示出AD相联紊乱疾病中NFT损害的早期改变，但却不能产生NFT和缺乏明显的神经病学症状。

（2）人类FTDP-17突变tau基因小鼠

目前已发现FTDP-17相联的tau基因三个突变位点P301L、V337M和G272V[75]，将人类FTDP-17突变tau基因导入小鼠中，进 行表达制作转基因小鼠。人类4Rtau基因同型体与突变型P301Ltau基因在鼠PrP启动子控制下，在JNPL3鼠系中得到表达[76]。该小鼠到 10月龄大时，90%的小鼠产生了运动和行为学紊乱，出现延迟的翻正反应并且最终不能进入直立状态。在悬挂实验中，小鼠短暂抓住绳子后就掉下，而转野生型 人tau的小鼠能用三个肢体和尾巴抓住保持不掉下。在出现症状两周以内，衰退扩展到所有的肢体，JNPL3小鼠几乎不能移动，出现张力障碍性体态。这些小 鼠体重减轻，梳理、发音和睁眼变得困难起来。而这些运动障碍使得这些小鼠不能在Morris游泳航行实验中进行测试。脑干、端脑和间脑中出现胶质细胞增 生，脊髓有48%的运动神经元减少。通过刚果红、Bielschowsky和Bodian银染方法在脑干、间脑、小脑和脊髓确认有NFT的形成。

6.其它的可用于制备转基因小鼠的基因 有报道其它的一些基因与AD的发病相关联，但大部分却是一个互相矛盾的数据。例如，有研究报道认为αI-抗凝乳蛋白酶与AD有关，而另外的研究却反对 [77]。对于极低密度脂蛋白受体、丁酰胆碱脂酶]和低密度脂蛋白受体]病例对照研究的结果也是互相冲突[78]。如果这两个与AD有关联的基因能导入 HLA-A2等位基因，将会给我们提供有用的信息。虽然TGF-β1在遗传学上与AD没有什么联系，但它在大脑损伤反应中扮演一个中心角色，小鼠大脑中的 星形胶质细胞增生反应时过度表达TGF-β1。hAPP和TGF-β1双转基因小鼠在2~3月时，大脑就出现了硫磺素S和Aβ免疫组化染色阳性，与大脑淀 粉样血管病变相似。而仅表达hAPP对照小鼠在2~3月时却没有观察到沉积，过度表达TGF-β1小鼠对hAPP的水平也没有影响[79]。如果新的与 AD发病有关的基因被确立，将会复制出具备更精确AD病理和症状的转基因动物。

转基因动物模型脑内出现了AD特有的病理改变Aβ沉积和SP，而FTDP-17突变tau基因小鼠出现了tau的病理改变，特别是NFT的形成，在AD动 物模型方面取得了巨大的进步，为针对SP和NFT方面的新药筛选提供了良好的动物模型。但建立转基因动物模型需要复杂的技术和高昂的费用，这类模型缺乏一 个衰老过程，不能完全具备理想AD动物模型特征，鼠系的遗传背景之间的差异导致同一类转基因鼠出现差异，重复性较差。因此，转基因模型要进一步推广应用， 尤其是在新药筛选和药理毒理学研究方面将受到很大局限。

四、以AD发病的环境因素为基础的动物模型

并不是所有的AD患者都具有阳性家族史，真正发生以上基因改变的患者所占的比例并不高，相当一部分患者从遗传学上是找不到任何线索，因此外界环境危险因素 在AD发病中也占有重要的地位。目前人们已发现一些环境因素能引起AD患者发病率升高。大体上包括生物因素（如病毒感染、高血压、糖尿病等）、物理因素 （如脑外伤等）和化学因素（如铝元素含量过高）。目前只有铝元素与AD发病的关系研究的比较多，并以此建立了铝元素慢性中毒的AD动物模型。

研究发现AD患者脑组织中铝含量增高，用原子吸收光谱仪测得铝在健康人脑中是1.8±0.8mg·g-1(脑组织干重)，AD患者脑内铝的含量是 3.6±2.9mg·g-1。而一般大于4mg·g-1即可引起神经细胞变性。游离铝可以自由通过嗅神经和血脑屏障进入脑组织的，在嗅球中含量最高，铝还 可以损害血管内皮细胞而破坏血脑屏障的完整性。铝进入脑内取代Ca2+和Mg2+，同氨基酸链上的谷氨酸和精氨酸结合形成谷氨酸铝盐或精氨酸铝盐的稳定复 合物，在大脑皮质、海马、杏仁核内富有谷氨酸神经元，铝多沉积于这些部位。铝在血液中可与转铁蛋白结合，在大脑皮质、海马、室中隔和颞叶杏仁处存在着许多 转铁蛋白受体，故铝易沉积于这些部位。动物实验结果表明，无论是家兔或大鼠经静脉、腹腔、皮下注射或是经胃肠道等途径给予铝化合物，均可见动物脑铝含量明 显增加，引起一系列行为异常，动物出现认知和记忆障碍的时间也早于其它指标。与幼年动物相比，铝的行为毒性多见于成年和老年动物，年龄越大，对铝的行为毒 性越敏感。铝的毒性引起脑内神经原纤维变性形成缠结已有许多报道。Deboni等给家兔皮下注射乳酸铝或酒石酸铝30天以上，可见脑干、大脑皮层及海马锥 体神经元中有NFT形成。其机制可能是铝能抑制蛋白磷酸酯酶2A和2B的活性，从而促使异常磷酸化的tau蛋白产生，继而导致NFT的形成。铝中毒可影响 脑内的神经递质的代谢，主要引起胆碱能神经元损伤，导致胆碱神经功能减退，引起学习记忆功能减退，对NE、5-HT、氨基酸及肽类等神经递质系统也有影 响。铝通过改变神经细胞的钙稳态，提高脂质过氧化损害细胞的作用，造成细胞结构和功能障碍，引起细胞的损伤和凋亡。铝引起脑内Aβ基因表达的增多也有报 道。钱亦华[80]等用AlCl3灌胃大鼠3个月建立老年性痴呆鼠模型，通过免疫组化的方法，发现其背海马结构内各部分均可见较多Aβ样免疫反应阳性神经 元，虽未有SP的形成，已说明铝引起Aβ在神经元内产生增多。铝引起Aβ在神经元内的沉积的机制可能与铝元素引起APP过度表达，Aβ过多的产生及铝元素 促进糖基与Aβ的定位连接有关。鉴于此，大多数学者认为铝在AD发病中起主要的作用，能够产生与AD相似的病理改变，形成铝元素中毒学说。 Mclachlan等还观察到三价金属螯合剂能减缓AD的病程[81]。但也有不少意见不同的报道，认为铝元素在AD发病中的并不是一个主要因素。而其它 的AD发病环境危险因素如何应用于动物模型，在动物身上研究其致病机理将是一个值得研究的课题。

五、多重复制AD模型

多重复制AD模型，是在现有模型的基础上，利用综合两种或两种以上模型方法获得的一种复合模型。这类模型可以综合获得各种已有模型特点，符合AD的多因素 发病机制。目前已有的动物模型有，将Aβ(1～40或25～35)和小剂量的IBO共同注入大鼠海马，2周后除出现神经元大量丢失外，注射点附近以及包括 CA1、CA2和部分脑回在内的脑区出现多数神经元异常。本实验室采用D-半乳糖和AlCl3合并制备的AD模型呈现了整体衰老、学习记忆功能下降、胆碱 能系统功能减退、脑组织β-APP、PS1、BACE基因表达明显增强，脑组织出现Aβ沉积并有类SP形成，成功地模拟了AD的发病及病理特征[82]。 在未发表的结果中，通过经典的Bieshowky银染的方法证实海马和大脑皮质神经纤维数量较正常对照组明显减少，排列紊乱和NFT形成。该方法造模简 单，造价低廉，适合大规模药物筛选。就现有的AD动物模型而言，由于没有一个完全模拟AD特征的动物模型，单一模型只模拟AD部分病理改变，把各种制作单 一模型的方法加以组合来模拟AD复杂的病因，是制备AD模型一个比较实际而有效的方法。

六、其它模拟AD部分特征的动物模型

冈田酸(okadaic acid，OA)损害模型：OA可以选择性抑制丝氨酸/苏氨酸蛋白磷酸酯酶1A和2A，引起大鼠脑内出现类似AD病理改变的双螺旋丝(paired helical filament，PHF)样的磷酸化tau蛋白和Aβ沉积[83]。

脑缺血痴呆动物模型：人脑的供氧系统对脑功能有着重要的保护作用，缺血缺氧首先影响的是脑的正常功能，此时极易出现智能障碍。通常老年动物脑血流量较成年动物减少20%以上，老年动物慢性脑缺血，可以产生类似AD的病理生理改变[84]。

β淀粉样肽(Aβ)损害模型：Aβ具有神经营养和神经毒性双重作用[85]。根据Hardy和Higgins 1992年在《Science》上提出的AD Aβ级联反应假说，Aβ沉积导致SP、NTF形成、细胞死亡和血管病变，最终导致痴呆。

第三节 小结和展望

目前AD实验动物模型的滞后在很大程度上制约了其治疗药物的筛选，还没有一个理想的AD治疗药物筛选模型。随着AD病因和发病机制知识的增加和更加完善的 AD动物模型的出现，我们将能够筛选得到具有良好开发前景的AD治疗药物，使这些药物经临床前客观评价后顺利进入临床试验。已有的AD动物模型，总的来 说，包括非转基因模型和转基因模型两大类。非转基因模型中，大部分AD模型只针对疾病某一方面的因素来制作AD模型，种类比较多，表现有空间学习记忆功能 的衰退，大多只针对AD病理的某一方面，相比AD错综复杂的病理过程来说有一定的差距。更重要的是缺乏AD脑内特征性变化，即AD病理改变中的SP和 NFT。非转基因模型的优点是其制备方法相对简单，重复性和稳定性较好，适合大规模药物筛选，其中多重复制AD模型能多更好的模拟出AD的多病因的综合作 用，基本具备了AD的病理特征，将是一个较好的药物筛选模型。

最近出现的转基因动物模型是AD模型中的一个热点，在小鼠中构建AD致病基因的转基因动物在AD动物模型方面前进了一大步，转基因小鼠在遗传学上模拟了 AD的发病，出现了许多AD的病理学特征的改变，包括细胞外Aβ的沉积、SP和NFT等，部分转基因小鼠还出现了炎性反应的改变。在这些小鼠模型中，Aβ 的浓度和Aβ沉积的大小能进行定量分析并且这些指标能反应出淀粉样沉积的严重程度。大脑Aβ水平的免疫学分析能被用来监测化合物的效果，并能进一步分析其 作用机制，是干扰APP到Aβ的代谢还是阻止Aβ的变性。使用AD转基因小鼠模型可以用于新药发现的领域，单一转基因小鼠通过杂交能产生转多基因的小鼠， 这些小鼠与转单基因的小鼠相比能较早的发展成Aβ沉积，可以加速实验的进程。因此，转基因小鼠对于评价一些针对Aβ和/或SP的产生、清除及对抗炎性反应 的药物是有用的。AD的潜在候选药物，如抗炎药物和β、γ分泌酶抑制剂，而后者能抑制蛋白水解酶的活性而阻断淀粉斑的形成从而减缓AD的发病进程。目前国 外已有β、γ分泌酶抑制剂的报道，由于γ分泌酶抑制剂只影响到Aβ产生对其他蛋白的代谢影响小，所以被认为是理想的作用靶点。γ分泌酶抑制剂按其核心结构 可分为为五类，分别为肽醛（peptide aldehydes）类、双氟酮(difluoro ketones)类、甲基丙烯酸类似物（hydroxyethyl analogues）、氨磺酰葑胺（fenchy lamine sulfonamide）类和其他类。它们采用的大都是体外培养的转APP基因细胞分析系统来检测抑制剂对APP代谢的影响。本实验室采用体外培养的转染 有AD患者APP及突变型PS1基因的CHO细胞系（M146L），使之产生Aβ42，建立体外AD治疗药物筛选模型，通过筛选治疗老年痴呆常用复方中使 用频率较高的一些单味中药及广东习用药材中的已知单体，初步发现广东海风藤中的五味子乙素具有明显抑制Aβ42产生的作用。表明五味子乙素很可能是一种潜 在的β和（或）γ分泌酶抑制剂，有可能成为治疗AD很有前景的药物，拟用转基因动物模型和多重复制模型进行体内的药效试验。目前，分泌酶抑制剂这类药物大 都尚处在动物实验室阶段，未见临床报道。

最近采用转基因动物作为AD模型的Aβ免疫治疗法，在动物试验上取得疗效后，已进入临床试验。Schenk等[86]用Aβ42免疫PDAPP小鼠发现能 抑制小鼠大脑中β淀粉样斑和相关的营养不良轴突的形成，紧接着小鼠Aβ42免疫法在几个实验室均取得成功。鉴于免疫治疗在转基因小鼠模型上取得的良好效 果，临床试验已经启动。在这些研究中使用的抗原，即AN-1792（Elan制药公司/Wyeth），是含有合成的Aβ42和佐剂（QS-21）的混合物。两个I期临床试验已完成，它们评价了AN-1792对AD病人的安全性和耐 受性。第一个是使用三个逐级提高的单独Aβ42剂量水平，佐剂的浓度是固定的。每一剂量组有8个病人，其提高剂量的间隔为6周。第二个I期临床试验是多剂 量（Aβ42和QS-21佐剂的剂量都在变）增加的研究，每个剂量水平之间的间隔为2~3月。

在这两个研究中，AN-1792具有良好的耐受性，Ⅱa期临床试验已经在美国和欧洲开始。Ⅱa期临床试验招募到375个AD病人，在主动治疗组 （n=300）已接受最高剂量Aβ42和最低剂量的在I期临床使用的QS-21。然而，这个试验由于出现了无菌性脑膜炎而被迫在2002年1月终止。所有 受影响的病人接受了1~3个剂量的AN-1792，并且他们的症状从最后一次接受药物注射后一直持续了5天到5个月不等。尽管AN-1792的没有继续进 行，但剩下的病人没有终止，并且这些病人将继续被监控其安全、中枢神经系统变化、认知状态和免疫功能1年。

这些副反应是没有预料到的，在5个不同种类的动物包括灵长类上所做毒性实验没有揭示脑炎和任何其它的副反应，在Ⅱa期临床试验开始时，接受I期临床实验 64个病人中没有一个产生了大脑炎症的症状。Ⅱa期临床试验中的一些病人出现的这个反应明显与免疫反应相联系，尽管其潜在的机制还没有找到。继续进行的测 试正在调查Ⅱa期临床试验中一些病人身上出现的反应是否与免疫反应相关，例如T细胞介导的针对AN-1792的免疫反应。虽然AN-1792没有继续进行 下去，由于引人注目的临床资料，人们还是在积极地追寻着安全、有效的免疫治疗方法。几个与AN-1792有很大不同的第二代免疫治疗产品正在活跃地研制之 中。

尽管疾病的一些重要标志物，如Aβ沉积和SP已在AD小鼠中出现，转基因小鼠也不能完全模拟出人类AD所观察到的所有特征。现在仍有争论是这些小鼠能否显 示出AD其它的一些明显而重要的特征，包括突触数量的减少，神经元的死亡或损失。这是由于AD的进程中可能包括各种不同病理途径的复杂的相互作用。APP 本身的过度表达产生的作用可能在转基因鼠中是一个模糊的部分，因为在一些APP基因过度表达的小鼠中缺乏神经元损失的特征，其原因可能是由于增加的 APP，同样导致潜在神经营养性APPs片段的增加。而产生的APPs-α，能阻止细胞的死亡。

即使已有的数据已表明鼠的大脑能重演AD的病理过程，但一些在AD转基因小鼠上发生的生物学效应并不同于人类的AD，例如，基因流行病学研究表明 ApoE4是AD的一个危险因子，但是在表达人类ApoE亚型而不含鼠ApoE的转基因鼠，早期Aβ沉积明显下降。另一个缺陷的

Hose Reel Cart

Hose Reel Cart,Portable Hose Reel Cart,Commercial Hose Reel Cart,Industrial Hose Reel Cart

NINGBO QIKAI ENVIRONMENTAL TECHNOLOGY CO.,LTD , https://www.hosereelqikai.com