Functional Roles of Amyloid-β Protein Precursor and ... - IOS Press

4 downloads 0 Views 1MB Size Report
[2] Pierce JE, Trojanowski JQ, Graham DI, Smith DH, McIntosh. TK (1996) ... Neve RL (1987) Amyloid beta protein gene: cDNA, mRNA distribution, and genetic ...
401

Journal of Alzheimer’s Disease 18 (2009) 401–412 DOI 10.3233/JAD-2009-1154 IOS Press

Mini-Forum Article

Functional Roles of Amyloid-β Protein Precursor and Amyloid-β Peptides: Evidence from Experimental Studies Mikko Hiltunena , Thomas van Groen b and Jukka Jolkkonen a,∗ a

b

Department of Neurology, University of Kuopio, Kuopio, Finland Department of Cell Biology, University of Alabama at Birmingham, Birmingham, AL, USA

Accepted 4 May 2009

Abstract. Amyloid-β (Aβ) has remained a central feature in research into Alzheimer’s disease (AD). Yet the function of the amyloid-β protein precursor (AβPP) and its processing products in the central nervous system is controversial. This review examines experimental literature from cell cultures to transgenic AD and brain injury models with a special emphasis on the functional role of AβPP and AβPP-derived peptides. AβPP knock-out mice exhibit severe metabolic abnormalities and behavioral deficits, indicating an important physiological function of AβPP. Also, an increasing body of evidence suggests that while Aβ is undoubtedly linked to neurodegeneration, the soluble amino-terminal fragment of AβPP (sAβPPα) has neuroprotective, neurotrophic, and cell adhesive functions. Moderate overexpression of human AβPP in rodents does not produce apparent Aβ pathology and has no significant effect on cognitive or sensorimotor behavior and, surprisingly, may even provide histological neuroprotection against focal cerebral ischemia. In contrast, phenotypes with more severe Aβ pathology show impaired cognitive performance, increased vulnerability to brain ischemia and trauma, and less favorable functional outcome even before Aβ deposition. A delicate balance in AβPP processing seems to determine its functional consequences. Thus, it is tempting to speculate that promotion of α-secretase-mediated cleavage of AβPP, which leads to increased sAβPPα production, provides a novel therapeutic strategy in the treatment of AD and brain injury. Keywords: Alzheimer’s disease, amyloid-β, amyloid-β protein precursor, brain trauma, cerebral ischemia, function, mice, rat

INTRODUCTION Amyloid-β protein precursor (AβPP) is a type 1 membrane protein with a long extracellular N-terminal and a short intracellular C-terminal domain. AβPP is sequentially processed by β- and γ-secretases to release amyloid-β (Aβ), a hydrophobic self-aggregating peptide consisting of 39 to 42 amino acid residues. Aβ 42 is ∗ Corresponding

author: Dr. Jukka Jolkkonen, Department of Neurology, University of Kuopio, Yliopistonranta 1 C, 70210 Kuopio, Finland. Tel.: +358 40 3552519; Fax: +358 17 162048; E-mail: [email protected].

a major component of senile plaques, one of the pathological hallmarks of Alzheimer’s disease (AD) [1]. In addition to AD, AβPP/Aβ is suggested to be strongly involved in the development of pathology after cerebral ischemia and brain trauma [2,3]. AβPP is a highly conserved protein between different species and is ubiquitously expressed in the central nervous system, indicating its important physiological function. Indeed, AβPP knock-out mice exhibit severe metabolic abnormalities and behavioral deficits [4], although functional redundancy among AβPP-like proteins (APLPs) is likely to partly compensate for the loss of essential AβPP gene functions. Supporting this notion, double knock-out mice (AβPP/APLP2) are

ISSN 1387-2877/09/$17.00  2009 – IOS Press and the authors. All rights reserved

402

M. Hiltunen et al. / Functional Role of AβPP Derived Peptides

Fig. 1. Schematic presentation of amyloid-β protein precursor (AβPP) and its processing products. AβPP undergoes sequential cleavage by α-, or β-, and γ-secretases leading to generation of the amyloid-β (Aβ) peptide or the p3 fragment. In addition, γ-secretase cleavage of AβPP C-terminal fragments (C83 and C99) at the ε-site results in the production of AβPP intracellular domain (AICD). Some important extracellular domains of AβPP are indicated, including Kunitz protease inhibitor (KPI) domain, Cu2+ and Zn2+ binding sites, and N - and O-glycosylation sites (CHO). The extracellular part of AβPP consists of several other important domains, such as neurotrophic (RERMS) and adhesion (RHDS) domains as well as glycosaminoglycan-, collagen- and laminin-binding domains (not shown). The Fe65 adaptor protein interacts with the AβPP intracellular domain (YENPTY motif) and it is needed for several AβPP-related functions in cell adhesion and neurogenesis.

not viable [5,6]. Functional diversity may also arise from different AβPP fragments, which are generated as a result of proteolytic cleavage of the full-length AβPP. While β- and γ-secretases produce the potentially toxic Aβ, a non-amyloidogenic pathway driven by α-secretase gives rise to soluble amino-terminal fragment of AβPP (sAβPPα), which has many neuroprotective, neurotrophic, and cell adhesive functions [7]. Recent data suggest that the C-terminus of AβPP is dispensable and that sAβPPα is sufficient to mediate the physiological functions of AβPP, such as spatial learning and long-term potentiation (LTP) [8]. Interestingly, BACE1-/- (β-site AβPP cleaving enzyme 1, i.e., β-secretase) mice bred to the PDAβPP line had severe behavioral impairments (e.g., sensorimotor impairments, spatial memory deficits, seizures) [9]. This also suggests that AβPP is necessary for certain specific physiological processes, although one should note that BACE1 has several additional substrates [10], which may play a role in behavior. Brain insults, such as stroke and trauma, lead to a variable upregulation and accumulation of AβPP/ Aβ [2,3]. The acute increase in AβPP expression represents most likely a neuroprotective response to brain insults [11], whereas long-term accumulation of Aβ peptides has negative consequences on behavior [12–14]. This review examines the current evidence from cell

cultures to transgenic AD mice and brain injury models to uncover the functional role of AβPP and AβPPderived peptides such as sAβPPα in the central nervous system. Regarding toxicity of Aβ and Aβ oligomers, readers are referred to the recent reviews [15,16].

NEUROPROTECTIVE, TROPHIC, AND ADHESIVE FUNCTIONS OF AβPP AND ITS PROCESSING PRODUCTS IN VITROAND IN VIVO AβPP undergoes sequential cleavage by α-, or β-, and γ-secretases leading eventually to generation of the Aβ peptide (amyloidogenic pathway) or the p3 fragment (non-amyloidogenic pathway) [17] (Fig. 1). In normal physiological conditions, AβPP is preferentially (∼90%) cleaved by α-secretase, which leads to release of sAβPPα. On the other hand, the membranebound AβPP C-terminal fragment known as C83 undergoes cleavage by γ-secretase, leading to formation of the non-amyloidogenic p3 fragment. The minor portion of AβPP (∼10%) is processed by β-secretase (BACE1), resulting in the formation of the secreted AβPPβ and an AβPP C-terminal fragment known as C99, which is consequently cleaved by γ-secretase producing Aβ. Apart from γ-secretase-mediated cleav-

M. Hiltunen et al. / Functional Role of AβPP Derived Peptides

age of both C83 and C99 at the γ-site, an additional γ-secretase cleavage site exists close to the cytoplasmic border (ε-site), which releases the AβPP intracellular domain (AICD). Subsequently, AICD can regulate transcription of genes by forming a transcriptionally active complex with the adaptor protein Fe65 and the histone acetyltransferase Tip60 [18]. The balance between amyloidogenic and non-amyloidogenic pathways is pivotal, since only small changes in this fine equilibrium may initiate pathogenic cascades ultimately leading to the onset of AD. Since the initial cloning of AβPP [19], cell culture and animal models have provided excellent tools to assess functional effects of AβPP and its processing products in different conditions. Attempts to reveal normal physiological functions of AβPP have systematically pointed towards neuroprotective, trophic, and adhesive nature of AβPP and its processing products. Several different functional domains have been identified in the extra- and intracellular parts of AβPP, including neurotrophic (RERMS) and adhesion (RHDS) domains, Kunitz protease inhibitor (KPI) domain, and extracellular glygosaminoglycan-, collagen- and laminin-binding domains [20–23] (Fig. 1). Furthermore, AβPP is comprised of extracellular metal binding components for copper and zinc, which have been suggested to control metal homeostasis in brain [24,25]. These domains and binding sites are part of the secreted sAβPPα peptide, which has been proven to exert the neuroprotective and trophic functions of AβPP. Although Aβ peptide is undoubtedly linked to neurodegeneration owing to its accumulation to amyloid plaques in the AD brain, neuronal functions of sAβPP were initially more elusive. Evidence from the fibroblast and primary neuronal cultures, however, emphasized the neuroprotective role of sAβPPα [7, 26]. sAβPPα was shown to exert its neuroprotective actions by increasing neuronal survival after hypoglycemic damage or glutamate-induced excitotoxicity, and also by preventing the rise of intracellular calcium ([Ca2+ ]i ) typically linked to these adverse conditions [7,27]. In addition, beneficial trophic properties of sAβPPα and AβPP were observed in the context of synaptic plasticity [7,28] and synaptogenesis [29, 30]. Importantly, neuroprotective properties associated with sAβPPα attenuated the Aβ-induced neurotoxicity as well as elevation of [Ca2+ ]i [27], which is a physiologically anticipated outcome considering that production of Aβ and sAβPPα from AβPP are mutually exclusive. Related to this, moderate overexpression of α-secretase, ADAM10 (a disintegrin and metallopro-

403

teinase) together with the overexpression of mutant human AβPP (AβPP V717I) in a bigenic mouse model (ADAM10-mo x AβPP V717I) resulted in reduced formation of Aβ peptides and almost completely prevented the plaque formation in the brain [31]. This effect, which was accompanied with the increased secretion of sAβPPα, associated functionally with the alleviation of impaired LTP and cognitive deficits. Recently, moderate overexpression of ADAM10 in mice was shown to increase cortical cholinergic, glutamatergic, and GABAergic presynaptic bouton densities, indicating that the endogenous sAβPPα has in vivo influence on neurotransmitter-specific cortical synaptic plasticity [32]. Taken together, these findings underline the possibility of modulating AβPP processing therapeutically in such a manner that α-secretase-mediated cleavage of AβPP is favored at the expense of Aβ generation in neurodegenerative diseases like AD. This approach has been used to develop therapeutic molecules, such as ladostigil, which possess diverse pharmacological properties and acts on several targets [33]. Ladostigil is a potential multifunctional drug targeted to AD and Lewy Body disease, which has been demonstrated to augment phosphorylated protein kinase C levels and consequently stimulate sAβPPα release in vitro. In terms of neuroprotection, the role of different AβPP isoforms is intriguing since the AβPP751 isoform or its sAβPPα derivative containing the KPI domain may have a greater neuroprotective capacity than the AβPP695 isoform lacking the KPI domain [11]. Related to this, cerebral ischemia leads to a transient upregulation of AβPP751 and 770 isoforms in the cortex adjacent to the boundary of ischemic lesion and white matter, suggesting an AβPP isoform-specific neuroprotection in response to neuronal injury [34–37]. It was also shown that overexpression of the human AβPP695 isoform provided neuroprotection after middle cerebral artery occlusion (MCAO) in rats [14]. The neuronspecific AβPP695 isoform has also been shown to play a role in learning and memory, since its expression was decreased in the rat hippocampal dentate gyrus during memory consolidation [37]. This suggests that AβPP participates, possibly in an isoform-specific fashion, in synaptic remodeling similar to other synaptic proteins, such as neural cell adhesion molecules. Moreover, AβPP C-terminal fragments appear dispensable in these processes, as sAβPPα alone is sufficient to mediate the physiological functions of AβPP, such as spatial learning and LTP in AβPP-deficient mice [38]. Low density lipoprotein receptor-like protein (LRP) is important for Aβ clearance in concert with apolipoprotein

404

M. Hiltunen et al. / Functional Role of AβPP Derived Peptides

E (ApoE) and α2-macroglobulin (α2M) in the blood brain barrier [39]. Interestingly, LRP is only able to bind sAβPP isoforms containing the KPI domain [40]. Related to this, it has been reported that sAβPPα containing the KPI domain inhibited the uptake of Aβ/α2M complex, while the uptake of Aβ/ApoE was unaffected [41]. Functional studies have recently provided novel data about the underlying molecular mechanisms of sAβPPα in neuronal growth and neuroprotection. A novel trophic function for sAβPPα was observed in organotypic hippocampal cultures, in which sAβPPα induced expression of several neuroprotective genes [42]. One of these sAβPPα-inducible genes was transthyretin, which was necessary for protection against Aβ-induced neuronal death according to the transthyretin antibody interference and knock-down experiments. Interestingly, chronic infusion of an antibody against transthyretin into hippocampus of transgenic mice overexpressing AβPP Swedish mutation resulted in exacerbated Aβ accumulation, tau phosphorylation, and neuronal loss in the CA1 region. The importance of transthyretin in neuroprotection was further emphasized by the observation that its overexpression protected AβPP23 AD mice from the behavioral and biochemical effects of Aβ toxicity [43]. Detailed molecular assessments revealed that transthyretin was neuroprotective, since it binds in a chaperone-like manner to aggregates of toxic and pretoxic Aβ in both the intracellular and extracellular environment. Consistent with these findings, Costa and co-workers [44] recently demonstrated that both recombinant transthyretin and transthyretin isolated from human sera were able to proteolytically process Aβ at multiple sites. Importantly, degradation of Aβ by transthyretin was inhibited by sAβPPα containing the KPI domain, while sAβPPα lacking the KPI domain did not affect degradation. The other significantly induced sAβPPα target gene was insulin-like growth factor-2 (IGF-2) [42], which has been shown to protect cells from Aβ-induced toxicity in vitro [45]. Collectively, the above-mentioned results highlight the downstream effects of sAβPPα in neuroprotective processes targeted to oppose the toxic effects of Aβ. Recently, sAβPPα was found to increase differentiation of human neural stem cells in vitro, while antibody interference of AβPP inhibited this effect [46]. Furthermore, transplantation of human neural stem cells into AβPP transgenic mouse brain led to glial differentiation instead of neuronal differentiation, suggesting that AβPP regulates neural stem cell biology in vivo.

Functional studies have also addressed the effects of sAβPPα on cell proliferation in the adult central nervous system [47]. It was shown that the subventricular zone of the lateral ventricle is a major sAβPPα binding site in the adult rodent brain and that the binding takes place in progenitor cells expressing the epidermal growth factor (EGF) receptor. Interestingly, blocking the secretion of sAβPPα or down-regulation of AβPP decreased the proliferation of EGF-responsive cells in vivo, resulting in reduction in progenitor cell pool. The fact that sAβPPα acts in synergy with EGF is interesting with respect to AβPP processing, as EGFR is one of the proposed target genes of AICD-mediated transcriptional regulation [48]. This may indicate a complex signaling system in progenitor cell proliferation, in which different processing products of AβPP may exert opposite effects on proliferation depending on the prevailing status of AβPP processing. Indeed, sAβPPα and AICD have been suggested to act in an opposite fashion in the TAG1-AβPP signaling pathway during neurogenesis [49]. TAG1 is a novel ligand for AβPP, which induces γ-secretase-mediated release of AICD and thereby negatively modulates neurogenesis. TAG1 suppression of neurogenesis is dependent on the Fe65 protein since wild-type AICD, but not AICD mutated at the Fe65 binding site, was able to reverse the aberrantly enhanced neurogenesis in homozygous TAG1 knockout mice. Apart from being an important determinant of neuroprotective and trophic processes, it has been evident for a long time that AβPP plays a role in cell adhesion and migration. The earliest evidence for AβPP involvement in cell adhesion came from the observation that the RHDS domain in AβPP promoted cell adhesion [21]. AβPP interacts with collagen and laminin via domains in its extracellular part, further supporting the role for AβPP in cell adhesion and neurite outgrowth [22,23]. The fact that AβPP co-localizes at the neuronal cell surface with patches of integrins [50] implied the possibility that AβPP is involved in intergrin-based cell adhesion and movement. Furthermore, AβPP and Fe65 were reported to be a part of the motility process, which involved AβPP and Fe65 interaction with β1-integrin at dynamic adhesion sites [51]. AβPP overexpression was shown to result in accelerated cell migration as assessed by the wound healing assay, and this effect was again further enhanced by co-expression of Fe65. The role of Fe65 in this process is likely linked to its well-established effect on the enhanced transport of AβPP to the plasma membrane. Consistent with these findings, blocking the binding of Fe65 to AβPP

M. Hiltunen et al. / Functional Role of AβPP Derived Peptides

slowed axonal outgrowth [52]. Recently, an elegant study employing a method of in utero electroporation of shRNA against AβPP into the developing rodent cortex revealed that AβPP is required for correct migration of neuronal precursor cells into the nascent cortical plate [53]. On the other hand, overexpression of cDNAs for AβPP or its homologues, APLP1 and APLP2, fully rescued the knock-down effect of AβPP on retarded cell migration. The role of full-length AβPP and sAβPPα in neurite outgrowth was further elucidated by in vitro and in vivo applications, which showed that sAβPPα regulates the function of AβPP in neurite outgrowth via competing with binding of full-length AβPP to β1-integrin [54]. Taken together, these data suggest that although sAβPPα encompasses independent functions, it also participates in the regulation of full-length AβPP function in specific cellular process, such as β1-integrin-based neurite outgrowth.

VARIABLE HISTOPATHOLOGY AND COGNITIVE IMPAIRMENT IN TRANSGENIC MOUSE MODELS OF ALZHEIMER’S DISEASE In AD, the two characteristic pathological markers of the disease are significant numbers of neurofibrillary tangles and neuritic plaques in the brain [55]. Neurofibrillary tangles consist of hyperphosphorylated, twisted filaments of the cytoskeletal protein tau [56], and are mainly found within cells, except for ghost tangles. The main components of plaques are extracellular Aβ peptides derived from AβPP as a result of proteolytic processing [57,58]. Most cases of AD are sporadic and have a late onset after 65 years of age. However, a small number of AD cases are of early-onset familial type and these cases are related to mutations in the genes for AβPP and presenilin 1 (PS1) and 2 (PS2) [59]. The mutations alter AβPP metabolism such that there is an increased production of Aβ, especially the longer and more fibrillogenic Aβ 42 [58,60]. Together these findings imply a central role for aberrant AβPP processing in the series of pathological changes occurring during AD, eventually characterized by the appearance of the typical neuritic plaques [1]. In the past decade, a number of genetically modified animal models for AD have been generated. While not every aspect of the transgenic mouse phenotype mimics that of human AD (e.g., neuronal loss, cholinergic deficit, and neurofibrillary tangles are not present in the mice), correlating Aβ and plaque deposition with progression of cognitive impairment provide a

405

valuable tool in understanding AD pathology and help in the search for new treatment strategies [61]. We have experience in several transgenic mouse models. One line of mice was generated from mating between AβPPswe and HuPS1-A246E single mutant mice (AβPP/PS1), which were originally produced at Johns Hopkins University [62], and was bred locally on a C57BL/6J background. The second line of AβPP/PS1 mice is the AβPPswe+PS1∆9 (AβPP/PS1∆), which was originally produced at Johns Hopkins University [63]. These mice are acquired from the Jackson Laboratory at the age of six weeks. The third line of transgenic mice is the AβPPswe/dutch/iowa (AβPP SweDI ), which has originally been produced at Stony Brook University [64]. Whereas the PS1 mutant mice never develop any visible pathology, all AβPP mutants develop pathology, although at different age. Furthermore, crossing AβPP mice with PS1 mutant mice accelerates Aβ pathology [62,65]. Whereas the single AβPP mutants develop pathology only after 18–24 months of age, the double mutants (AβPP/PS1) develop deposits after about 6 months of age [62]. The distribution of the pathology in the three transgenic AD mouse lines is also distinct. Both AβPP/PS1 lines have predominantly parenchymal deposits of Aβ with varying levels of diffuse deposits (Fig. 2). Whereas the AβPPSweDI also develop significant Aβ deposits in blood vessel walls (cerebral amyloid angiopathy) in the parenchyma, AβPP/PS1 mice predominantly exhibit Aβ deposits in pial vessels and penetrating arterioles. Plaques are always surrounded by activated astrocytes and microglial cells, whereas diffuse deposits are not [66]. Similarly, Aβ deposits in blood vessel walls are not associated with signs of inflammation. It is of interest to note that the two AβPP/PS1 lines do not have many plaques or Aβ deposits in the white matter. The AβPPSweDI line, however, does develop white matter pathology. More specifically, fiber tracts show axonal varicosities that are positive for Aβ and AβPP. All of these mouse models of AD are characterized by cognitive impairments with age, which are correlated with the amount of Aβ present in the brain [67– 70]. In some mouse lines it has been demonstrated that differences in LTP are also present [68,71], further indicating that overexpression of AβPP can lead to changes in synaptic transmission and cognitive impairment. Again this seems to depend on severity of Aβ pathology, since in transgenic mice overproducing AβPP with no sign of Aβ plaques, synaptic plasticity is enhanced [72]. In contrast to AD, most transgenic models do not show any loss of neurons, except the AβPP23 mice, in which some neuronal cell loss in the CA1 area of the hippocampus is observed [73].

406

M. Hiltunen et al. / Functional Role of AβPP Derived Peptides

Fig. 2. Photomicrographs showing amyloid-β (Aβ) and amyloid-β protein precursor (AβPP) pathology in three transgenic mouse lines of Alzheimer’s disease. The pathology in the two APP/PS1 lines is similar, though the APP/PS1∆ line has a much faster time course. The APPSweDI line has predominant blood vessel depositions. Bottom row shows higher power images of plaques stained with antibodies against AβPP C-terminus, showing labeled dystrophic neurites. Most of AβPP is present surrounding the plaque core. Data modified from van Groen et al. [66].

NEUROPROTECTION IN hAβPP RATS AFTER FOCAL CEREBRAL ISCHEMIA IS NOT TRANSLATED TO FUNCTIONAL OUTCOME Increasing evidence suggests a relationship between cerebral ischemia, amyloid pathology, and AD [74– 76]. In experimental stroke models, transient upregulation and accumulation of AβPP are detected in cortical areas adjacent to the infarct and in the subcortical white matter [34,35,77,78]. We have confirmed this in MCAO rats [3]. Namely, shortly after the MCAO there is a significant amount of AβPP, as labeled with an antibody against its C-terminus, in axonal bulbs in the white matter both near the infarct and remote from the infarct. Furthermore, there is AβPP labeling in the endings of the damaged axons, especially in the thalamus (Fig. 3, black arrows). It should be noted that at short time intervals of less than 1 week after the MCAO, AβPP is present in degenerating axon terminals, but at longer survival times, only extracellular deposits are present (Fig. 3). In the thalamus, AβPP is still present even after several months of survival. Furthermore, it should be noted that at 1 week and 1 month after MCAO, AβPP expression is significantly increased in

Fig. 3. Six photomicrographs showing amyloid-β protein precursor (AβPP) pathology in the thalamus following transient middle cerebral artery occlusion (MCAO) in rats. AβPP, as assessed by an antibody against its C-terminus, is present both in axon endings (black arrow) and intraneuronally (gray arrow) 1 week after MCAO, in extracellular diffuse clumps with fewer labeled neurons 1 month after MCAO, and in extracellular plaque-like deposits 10 months after MCAO. Please note that after 1 month only the soma of the neurons is labeled, compared to both soma and dendrites after 1 week of survival. Data modified from van Groen et al [3].

neurons surrounding the AβPP-labeled axon endings (Fig. 3, gray arrow). Similarly, in the white matter the AβPP deposits persist for at least up to 6 months after MCAO. Extracellular Aβ (plaques) is likely to inter-

M. Hiltunen et al. / Functional Role of AβPP Derived Peptides

fere locally by taking up space and displacing synapses and dendrites, but axonal AβPP accumulation interferes with axonal transport, likely leading to impaired synaptic function and cognitive deficit. Whether AβPP overexpression exacerbates ischemic damage has been studied in transgenic mice. Permanent middle cerebral artery occlusion in AβPP695swe (Swedish mutation) mice showed increased susceptibility to ischemic injury with no change in Aβ 40 or Aβ42 levels [79]. The effect was related to a loss in endothelium-dependent vascular reactivity and more severe ischemia in regions at risk for infarction. Cerebrovascular dysfunction was also present in both young and old Tg2576 mice [80]. Another study using AβPP751 transgenic mice showed increased ischemic vulnerability associated with enhanced microglia activation and inflammation [81]. Using a novel line of transgenic hAβPP695 rats, we have assessed the histological and functional consequences of hAβPP upregulation after transient MCAO [13,82]. The rats were generated to overexpress the human isoform of AβPP695, but have not been found to develop senile plaques typical of AD. Biochemical analyses in adult, intact hAβPP695 transgenic rats showed that total Aβ 42 levels in the cortex and hippocampus were twice as high as those found in wild-type controls (Fig. 4A). MCAO produced significantly smaller lesions in the cortex in hAβPP695 transgenic animals than in wild-type animals (Fig. 4B). Consistent with this, neuronal overexpression of hAβPP695 or hAβPP751 at moderate levels protects neurons against kainate-induced neurotoxicity [11]. Interestingly, transgenic hAβPP695 animals performed significantly worse on the beam-traversing task as compared to the wild-type rats following MCAO following MCAO [14]. While there were no differences between the groups prior to surgery, all ischemic animals showed impaired hindlimb function following MCAO. In addition, behavioral performance was similar at 3 weeks post-surgery among ischemic rats, but thereafter at 9 and 15 weeks hAβPP-MCAO animals made significantly more foot faults than the wildtype rats after MCAO (Fig. 4C). Similarly, hAβPP695 rats showed an acquisition (learning) impairment compared to sham-operated controls in the water maze at 16 weeks post-surgery. These seemingly incongruent results may be due to the chronic accumulation of Aβ peptides in the brain parenchyma and particularly in the thalamus, which subsequently impairs neuronal function and has a negative impact on behavior. There are not many in vivo studies on the effect of AβPP/Aβ fragments in cerebral ischemia. Whitehead

407

and co-workers [12,83] studied the interaction between Aβ25−35 -induced toxicity and cerebral ischemia. A combination of Aβ 25−35 and brain ischemia exacerbated inflammatory responses in the hippocampus and decreased performance in the Montoya’s staircase test measuring forelimb skilled reaching ability [12]. Also, a time-dependent increase in the infarct size and memory and learning deficits was observed [83]. Thus, it is apparent that cerebral ischemia in the presence of high levels of amyloid fragments leads to impaired functional outcome through neuroinflammation. Another study examined the effect of intracerebral injection of Aβ oligomers in the four-vessel transient cerebral ischemia model [84]. The data revealed that Aβ oligomers caused impaired spatial memory and decreased acetylcholine release in the CA1 area, but no apoptotic cell death. It seems that the immediate increase in AβPP expression following cerebral ischemia may be part of the cellular response to stress and serve neurotrophic and protective functions [85,86]. Although initially Aβ may depress excitatory synaptic neurotransmission [87], in the long run ischemia-induced cleavage of AβPP into the toxic Aβ fragments [88] disrupts neuronal function and increases cell death through multiple mechanisms, including impaired calcium homeostasis [7], initiation of inflammatory processes [12,89], and loss of vascular reactivity [79].

AXONAL ACCUMULATION OF AβPP/Aβ FOLLOWING BRAIN TRAUMA AND ITS BEHAVIOURAL CONSEQUENCES Traumatic brain injury (TBI) is the leading cause of death and the most common cause of permanent disability in young adults [90]. It also increases the susceptibility to AD [91,92]. TBI in rats is characterized by enduring cognitive, sensorimotor, and histopathological changes persisting up to one year after injury [93]. Histopathology includes a progressive expansion of the cortical cavity and axonal swelling in the cortex, striatum, white matter, and thalamus. Interestingly, AβPP and AβPP-like proteins accumulate in swollen axons following TBI, being one of the earliest markers of axonal injury [2]. The axonal injury may cause disconnection between cortical and thalamic neurons and lead to delayed apoptotic death [94] or neuronal atrophy [95]. Tau protein has an important role in initiating this pathological cascade. Altered tau phosphorylation or tau levels destabilize the microtubule net-

408

M. Hiltunen et al. / Functional Role of AβPP Derived Peptides

A)

Ab1-42

100

B) *

90

Infarct volumes

90

= WT = hAbPP

80

80

**

60

70

mm3

70

pg/ml

100

50 40

60 40

30

30

20

20

10

10

0

0

Cortex Hippocampus

*

50

Cortex

Striatum

Beam-walking test/hindlimb

C)

Slip ratio (% )

80 70

**

*

60

= SHAM = WT+MCAO

50 40 30

= hAbPP+MCAO

20 10

ee k

21

15 W

3

9

W ee k

W ee k

W ee k

Ba

se l

in e

0

Fig. 4. A) Increased Aβ42 concentration in the cortex and hippocampus of rats overexpressing human amyloid-β protein precursor (hAβPP) (*p