Seeking The Prime Mover In Animal Models

The human sequence of the A/ peptide is particularly prone to aggregation in vitro and is required for deposition in vivo. The murine sequence contains three amino acid substitutions, F4G, Y10F, H13R, all in the hydrophilic N-terminus of the peptide (Figure 11.2), resulting in greatly reduced fibril-forming activity, which may explain why rats and mice don't spontaneously develop plaques. Initial attempts at modeling amyloid deposition involved infusion of the human sequence peptide or plaque cores into the brains of recipient animals. In rat models, the infused peptide itself can form deposits in brain, but these lesions lack many features of bona fide plaques, and, unlike AD, the deposition is not sustained by endogenous processes. In a monkey model of synthetic A/ fibril infusion (Geula et al., 1998), some neurodegeneration at the site of injection was observed, but amyloid plaque pathology did not develop. Similar results were obtained in rodent brain (Frautschy et al., 1996).

A/ amyloid pathology has been successfully induced in animals by taking a lead from the prion field, where mouse models to study infectious prions were created in which the species barrier had been circumvented by overexpressing the cognate normal prion protein in mice. Seeding of new /-amyloid plaques in the parenchyma and the walls of the cerebrovasculature was engendered in Tg2576 animals with high interstitial fluid levels of A/ by a single injection of dilute AD brain extract (Walker et al., 2002). A/ proteopathy was not inducible in age-matched, nontransgenic mice, and the injection of brain extracts from young subjects was ineffective. To date, injections of pure, synthetic A/(1-40) and A/(1-42) have failed to instigate /-amyloidogenesis in the same way as brain extracts, suggesting that there is something special about the A/ in the extract that is not found in the synthetic material. Identifying the properties of the brain extracts that promote /-amyloidogenesis will yield important information about the events that initiate this process in vivo.

Transgenic expression models Success in obtaining plaque deposition in transgenic mice initially required overexpression of the human /APP relative to the endogenous mouse protein. Transgenic animals with human /APP-overexpression under the control of neuron-selective or neuron-specific promoters deposit A/ peptides in parenchymal plaques in many of the same areas as in the AD brain as well as around cerebral and leptomeningeal blood vessels. Deposition is age- and protein level-dependent, usually beginning at 6-10 months in the absence of accelerating mutations. Dystrophic neurites develop around condensed senile plaques, as do reactive microgliosis and astrogliosis. While phosphorylated tau epitopes become prominent in dystrophic neurites, true neurofibrillary tangles never develop, and large-scale neuronal cell death has not been documented. A number of ^APP transgenic models have been produced that employ a variety of promoters, expression levels, and ^APP splicing variants. The upshot is that human ^APP overexpression in mice reproduces CNS ^-amyloidosis and (to a variable degree) amyloid angiopathy in specific brain regions, but the mice do not develop the full histopathology of AD. ^APP overexpression also produces deficits in synaptic physiology (particularly in long-term potentiation [LTP]) and in several behavioral paradigms thought to reflect memory function in humans well before A^ deposits are his-tologically demonstrable. While potentially due to the overexpression of ^APP, which is believed to mediate cell-cell contact and signaling, these same effects can be induced by application of soluble oligomeric forms of A^, and the deficits are abolished by antibodies to A^ peptide. Only some of the most widely used or most informative models are discussed here.

Caveats. In evaluating the histopathologic effects of murine transgene expression, it is important to recognize the presence of nonspecific lesions that develop with age in various strains of mice. Specifically, clusters of inclusions emerge most prominently in the hippocampus but to a lesser degree also in the cerebellum and other areas (Jucker et al., 1992). These inclusions (sometimes referred to as Jucker Bodies) occur within astrocytes and do not closely resemble any known human lesions. The inclusions can react nonspecifically with many antibodies, particularly polyclonal antibodies, and their presence is governed by a variety of factors, including the age, strain, and possibly sex of the mice. It is imperative to distinguish these inclusions from transgene-related neuropathology.

It is also important to recognize that the gender of the mice can influence the degree of A^ pathology. In Tg2576 ^APP-transgenic mice, amyloid deposition is significantly greater in females than in males (Callahan et al., 2001).

PDAPP. The first compelling report of a transgenic mouse that deposited A^ plaques used ^APP V717F familial AD mutation minigene construct containing some intronic sequences under the control of the platelet-derived growth factor-^ (PDGF-^) promoter (Games et al., 1995). The V717F mutation, which is near the gamma-secretase cleavage site, promotes the selective production of the highly amyloidogenic A^42 peptide. Plaque deposition is noticeable around 6 months of age. Between 6 and 12 months plaque deposition increases, and virtually all plaques up through this age are neuritic and Thioflavine S-positive. After 12 months, A^-deposition increases precipitously, much of which is diffuse and Thioflavine S-negative. The dentate gyrus and entorhinal cortex are heavily invested (Reilly et al., 2003). There is no change in hippocampal volume from

3 to 22 months in the transgenic animals compared to nontransgenic littermates. Four to five months before plaque deposition there are significant differences in electrophysiology and behavior in the transgenic animals. This model has been a workhorse driving anti-A^-based therapeutic development for Elan (Athena) and Lilly, but PDAPP mice were not made generally available to the research community. At Elan, they were used for the studies that first demonstrated that immunization with the A^ peptide would prevent or clear brain A^ deposits, currently a promising therapeutic strategy for AD and an area of active research.

Tg2576 - Hsiao mouse. The human ^APP695 amino acid protein with a Swedish K670N/M671L mutation coupled to a mouse prion promoter was designed to express the human protein in neuronal cells (Hsiao et al., 1996), although this promoter seems to be active in other cell types such as vascular smooth muscle cells. A five- to six-fold overexpression was achieved in this early model, which is sufficient to drive deposition by 9 months of age, increasing dramatically up to 30 months. The Swedish mutation increases only ^-site cleavage; as a result, there is a higher ratio of A^40:A^42 produced, and the plaques are a mixture of diffuse Thioflavin S-negative and neuritic Thioflavin S-positive deposits. Neuritic plaques are associated with abnormal, tau-immunoreactive neurites and with a glial inflammatory response. An electrophysiological and behavioral pheno-type that is typical of the ^APP overexpressing mouse models is present in these animals. These deficits are established before plaque deposition is apparent. Maintenance on a SJL/BL6 hybrid strain background and ^APP heterozygosity are required because of ^APP toxicity and of the site of insertion of the transgene array, respectively. These mice were made available both to industry and academia, and as a result, Tg2576 has become the most thoroughly studied of the murine ^APP models. They are now commercially available from Taconic.

Caveat. Several inbred mouse strains, including SJL, carry the rd (retinal degeneration) mutation, a recessive, null mutation in the ^-subunit of rod-specific cyclic GMP phosphodiesterase. Homozygous rd animals lose their rod cells between postnatal days 8 and 20, followed by a more gradual disappearance of the cone cells, such that by

4 weeks of age, the retina is devoid of photoreceptors. Obviously, this condition can profoundly influence performance on vision-dependent behavioral tasks.

APP23 mice. These mice express ^APP751 with a Swedish K670N/M671L mutation under control of the neuron-specific Thy 1 promoter. The 751 isoform of ^APP is not normally expressed in neurons, which produce mostly ßAPP695. Plaque deposition begins around 6 months of age (0.3% of the cortical area) rising to 9% area at 22 months. Electrophysiological and behavioral deficits are noted before the onset of deposition, and condensed plaques induce inflammatory responses in the surrounding glial elements. Phospho-tau epitope induction is documented in neurites, but as in other ßAPP-transgenics, there are no neurofibrillary tangles. Some neurodegeneration is also reported, which is unusual in ßAPP models, but neocortical synapses are not depleted even when amyloid load is high (Boncristiano et al., 2005). APP23 mice have a particularly strong propensity to develop Aß accumulation in the vascular wall, and thus are a useful model of cerebral amyloid angiopathy (CAA) (Vloeberghs et al., 2004).

CRND8 mouse. The double ßAPP695 Swedish K670N/M671L-Indiana V717F mutant under control of the hamster prion protein promoter produces a massive amount of Aß and a particularly virulent ß-amyloidosis. The Swedish mutation increases the total amount of peptide produced by altering the ßAPP sequence immediately N-terminal to the ß-secretase cleavage site, making it a much better substrate for BACE. The Indiana mutation alters the ßAPP sequence immediately C-terminal to the y-secretase cleavage site causing more Aß(1-42) than Aß(1-40) to be present after 8 weeks of age. The CRND8 mouse attains significant Thioflavine S-positive Aß deposition at 3 months of age in 100% of the animals, and both diffuse and neuritic plaques are apparent by 6 months. Neurofibrillary tangles are not present, but studies of the phospho-tau epitopes in neurites, which are seen in all other ßAPP models, have not been published. The lack of neuronal cell death with the high level of ßAPP expression achieved in these animals is unusual. A possible explanation for this is the C3H genetic background, which is resistant to ßAPP toxicity. Behavioral and electrophysiological effects are noted early and can be reversed by immunization with anti-Aß antibodies.

Hereditary cerebral Aß angiopathy. Several ßAPP familial mutations are associated with pathologies other than pure AD. ßAPP mutations that change specific amino acids within Aß (at positions 21-23) cause rare, autosomal dominant Aß-proteopathies characterized by profuse CAA (Revesz et al., 2003). The Aß(1-40) peptide is the main deposited species in these mutants; as deposition progresses, the smooth muscle cells of the tunica media are lost, and the vessel walls lose their elasticity and become susceptible to rupture. While hereditary CAA is rare, wild-type Aß is deposited in the cerebrovasculature of nearly all AD subjects as well as Down Syndrome patients, increasing the probability of hemorrhage.

Until recently, CAA has been difficult to model in mice bearing the human familial mutations. Mice transgenic for ßAPP770 Swedish K670N/M671L + Dutch E22Q and

Iowa D23N under control of the Thy 1.2 promoter express very low levels of /APP770, less than the endogenous mouse /APP (Davis et al., 2004). Nevertheless, the mice develop significant accumulations of mostly fibrillar vascular and perivascular A/ by 3 months of age. After 6 months, about 50% of vessels have deposits, and diffuse plaques appear in brain forebrain parenchyma. By one year of age there is robust deposition of A/ in microvessels and within the brain parenchyma. Part of the reason for the accumulation of the (Dutch, Iowa) A/(1-40) is that, at physiologic concentration, the mutant peptides are 10-fold less readily cleared than the wild-type peptide.

In another model of CAA, neuronal overexpression of Dutch mutant E693Q h/APP driven by the Thy-1 promoter produced a phenotype characterized predominantly by CAA, whereas overexpression of wild-type /APP caused mainly parenchymal (plaque) amyloidosis (Herzig et al., 2004). Interestingly, when the Dutch mutant mice are made doubly transgenic by crossing with mutant presenilin-1-transgenic mice, the most abundant form of brain A/ is changed from A/40 to A/42, and the amyloid pathology is shifted from the vasculature to the brain parenchyma.

Double transgenics In an effort to produce animals that more rapidly develop plaque pathology and a more severe phenotype, presenilin (PS) mutants were crossed with h/APP-expressing animals. A/ deposition requires the presence of the human A/ peptide, as transgenic mice expressing only mutant PS fail to develop plaque pathology or CAA. A number of different PS1 mutant double transgenics reproduce the pattern of elevated A/42/40 ratios seen in the human familial AD cases (Borchelt et al., 1996; Duff et al., 1996; Citron et al., 1997). Plaque formation in the transgenic animals mimics the aggressiveness of the familial disease as judged by the age of onset and rate of progression of the human cognitive symptoms. Plaques are detected as early as three months of age in the double mutants. There are some differences depending on which PS1 mutant is used.

bAPPSWE x PS1(M146L) (Duff mouse). This mouse was constructed by crossing the human /APPSWE Tg2576 (Hsiao mouse section) with human PS1 (A246E) mutant mice (Holcomb et al., 1998). Since the transgenes are both under the control of the mouse prion protein promoter, they are expressed in the same cells.

bAPPSWE x PS1(A246E) (Borchelt Mouse). These mice overexpress murine /APP695 in which the mouse /APP A/ region is replaced with the human A/ sequence with the Swedish K670N/M671L mutation (Borchelt et al., 1997). The mice were crossed with human PS1 (A246E) mutant mice, and both transgenes are under control of the mouse prion promotor. A 50% increase in the ratio of A/42/A/40 is observed in the brains of these animals. A^ accumulation is accelerated by several months, although the eventual degree of deposition is similar in the APPSWE and the APPSWE/PS1 animals.

Caveat. While useful in screening for agents that will prevent amyloid deposition, transgenic mice expressing mutant PS1 come with complications for studying mechanisms of neurodegeneration. The sporadic form of AD does not involve PS1 mutations or ^APP mutations. PS1 is thought to be the catalytic component of the intramembrane y-secretase complex, which includes a number of proteins of unknown function. Mutations in PS1 alter the processing of other y-secretase substrates besides ^APP, such as Notch as well as unknown substrates. There is one PS1 mutant (out of over 150 different mutations) that demonstrates frontotemporal dementia-like neurodegeneration in the absence of changes in ^APP metabolism (Dermaut et al., 2004). Thus, mutations in PS1 could have effects on cellular physiology beyond ^APP.

Even with both transgenes present, the amount of neurodegeneration observed in ^APP/PS1 animals is disappointingly meager compared to that occurring in AD brain. The reason for the lack of extensive neuronal death is unknown. Possible explanations include an incomplete inflammatory response resulting from a deficiency in complement components in many mouse strains, augmented metabolism or sequestration of intracellular A^, or a missing linkage between PKR and JNK/p38 stress kinase systems in mice (Peel and Bredesen, 2003).

Non-overexpressing transgenic models Early experiences in trying to create mice that would produce A^ plaques appeared to dictate that levels of ^APP 5-6 times the endogenous mouse ^APP were required for deposition. Several laboratories produced models that were expected to be more like the sporadic AD situation, making do with less expression and leaving the ^APP gene in its normal chromosomal context.

bAPP Knock-in mice. Murine ^APP was modified by inserting the Swedish mutation immediately N-terminal to the ^-secretase cleavage site and humanizing the three residues in the mouse A^ sequence that differ from those in humans (H:M = R5G, Y10F, H13R) (Siman et al., 2000). The hybrid full-length £APP was under the control of the endogenous mouse promoter and retained the native mouse introns and splicing signals. In these animals humanized ^APP is produced in the absence of the mouse peptide in homozygotes at a normal level. The Swedish mutation causes a 9-fold increase in the amount of A^ present as a result of its enhancement of ^-secretase cleavage. These mice produce human sequence A^ in the absence of the mouse sequence protein and develop plaques around 20 months of age. Other characteristics of the pathology are the same as in the overexpressors.

YAC mice. Another approach designed to maintain appropriate promoter control and splice variant production employed inserting the genomic sequence for the entire human ^APP gene, including introns and exons for the 770 residue protein (~400kb) plus flanking sequences (~250 kb) in a yeast artificial chromosome construct (Lamb et al., 1999). Expression levels are modest, 2-3 times the endogenous mouse protein, and the tissue-specific ratios of splice variants are retained. The Swedish mutation (K670N, M671L) increases the amount of A^ peptides produced while (V717I) increases the proportion of A^(x-42). The Swedish mutation animals produce plaques at 14 months but with a different anatomical distribution than the Tg2576 cDNA-equivalent overexpressing mice. Most transgenic ^APP models employ neuron-specific/selective promoters. The results with the native ^APP promoter and introns suggest that ^APP splicing to produce isoforms and/or promoter activity in different brain regions or cell types such as glia can influence the development of pathology. This is a particularly important concept to consider when employing transgenic models to study mechanisms.

Blood Pressure Health

Blood Pressure Health

Your heart pumps blood throughout your body using a network of tubing called arteries and capillaries which return the blood back to your heart via your veins. Blood pressure is the force of the blood pushing against the walls of your arteries as your heart beats.Learn more...

Get My Free Ebook


Post a comment