Genetic and molecular aspects of Alzheimer's disease shed light on new mechanisms of transcriptional regulation.

Rapid advances made in biological research aimed at understanding the molecular basis of the pathogenesis of Alzheimer's disease have led to the characterization of a novel catalytic activity termed gamma-secretase. First described for its beta-amyloid-producing function, gamma-secretase is now actively studied for its role in a novel signal transduction paradigm, which implicates cell-surface receptor proteolysis and direct surface-to-nucleus signal transduction. gamma-Secretase targets numerous type I protein receptors involved in diverse functions ranging from normal development to neurodegeneration. In this Review we discuss how the study of the genetic and molecular aspects of Alzheimer's disease has revealed a dual role of gamma-secretase in transcriptional regulation and in the pathogenesis of familial Alzheimer's disease.

Stereologic assessment of the total cortical volume occupied by amyloid deposits and its relationship with cognitive status in aging and Alzheimer's disease.

Although the presence of amyloid deposits is required to establish the neuropathologic diagnosis of Alzheimer's disease, from a clinical point of view, a direct contribution of these cerebral lesions to cognitive deficits is still controversial. The development and standardization of quantitative and accurate biochemical and neuropathologic methods may be critical to improve the postmortem diagnosis and clinicopathologic correlations. Here, we used a point counting method, based on the Cavalieri principle, to estimate the volume occupied by amyloid deposits in a discrete region of the prefrontal cortex and in the hippocampal formation, in brains from patients with cognitive status ranging from normal to severely demented. We demonstrate that the assessment of the total volume occupied by the amyloid deposits in the entorhinal cortex and subiculum can be considered an effective predictor of dementia severity. We also reveal the existence of a high degree of regional and interindividual heterogeneity in amyloid distribution and relative volume. Our data suggest that even though a correlation was observed between the stereologic point counting method and a non-stereologic random field thresholding approach, in most cases non-stereologic methods may not provide adequate samples of the tissue and may lead to unreliable estimates of amyloid burden due to the inhomogeneous distribution of amyloid in the cerebral cortex and the large variability among brains.

Appican, the proteoglycan form of the amyloid precursor protein, contains chondroitin sulfate E in the repeating disaccharide region and 4-O-sulfated galactose in the linkage region.

Chondroitin sulfate (CS)-D and CS-E, which are characterized by oversulfated disaccharide units, have been shown to regulate neuronal adhesion, cell migration, and neurite outgrowth. CS proteoglycans (CSPGs) consist of a core protein to which one or more CS chains are attached via a serine residue. Although several brain CSPGs, including mouse DSD-1-PG/phosphacan, have been found to contain the oversulfated D disaccharide motif, no brain CSPG has been reported to contain the oversulfated E motif. Here we analyzed the CS chain of appican, the CSPG form of the Alzheimer's amyloid precursor protein. Appican is expressed almost exclusively by astrocytes and has been reported to have brain- and astrocyte-specific functions including stimulation of both neural cell adhesion and neurite outgrowth. The present findings show that the CS chain of appican has a molecular mass of 25-50 kDa. This chain contains a significant fraction (14.3%) of the oversulfated E motif GlcUA beta 1-3GalNAc(4,6-O-disulfate). The rest of the chain consists of GlcUA beta 1-3GalNAc(4-O-sulfate) (81.2%) and minor fractions of GlcUA beta 1-3GalNAc and GlcUA beta 1-3GalNAc(6-O-sulfate). We also show that the CS chain of appican contains in its linkage region the 4-O-sulfated Gal structure. Thus, appican is the first example of a specific brain CSPG that contains the E disaccharide unit in its sugar backbone and the 4-O-sulfated Gal residue in its linkage region. The presence of the E unit is consistent with and may explain the neurotrophic activities of appican.

Presenilin-1 binds cytoplasmic epithelial cadherin, inhibits cadherin/p120 association, and regulates stability and function of the cadherin/catenin adhesion complex.

Here we show that presenilin-1 (PS1), a protein involved in Alzheimer's disease, binds directly to epithelial cadherin (E-cadherin). This binding is mediated by the large cytoplasmic loop of PS1 and requires the membrane-proximal cytoplasmic sequence 604-615 of mature E-cadherin. This sequence is also required for E-cadherin binding of protein p120, a known regulator of cadherin-mediated cell adhesion. Using wild-type and PS1 knockout cells, we found that increasing PS1 levels suppresses p120/E-cadherin binding, and increasing p120 levels suppresses PS1/E-cadherin binding. Thus PS1 and p120 bind to and mutually compete for cellular E-cadherin. Furthermore, PS1 stimulates E-cadherin binding to beta- and gamma-catenin, promotes cytoskeletal association of the cadherin/catenin complexes, and increases Ca(2+)-dependent cell-cell aggregation. Remarkably, PS1 familial Alzheimer disease mutant DeltaE9 increased neither the levels of cadherin/catenin complexes nor cell aggregation, suggesting that this familial Alzheimer disease mutation interferes with cadherin-based cell-cell adhesion. These data identify PS1 as an E-cadherin-binding protein and a regulator of E-cadherin function in vivo.

Amyloid beta peptides activate the phosphoinositide signaling pathway in oocytes expressing rat brain RNA.

Amyloid beta peptides (Abetas) of 39-43 amino acids constitute the major protein component of the amyloid plaques found in Alzheimer's disease brain. The generation of Abetas is regulated by the phosphoinositide (PI) pathway, which commonly couples to transmitter receptors. This study reports evidence for the activation of the PI pathway by Abetas in Xenopus oocytes expressing rat brain RNA. The naturally occurring peptides Abeta1-40 and Abeta1-42 were both active, whereas the cytotoxic fragment Abeta25-35 and the reverse peptide Abeta40-1 did not stimulate the PI pathway. Abetas rapidly lost potency in solution, suggesting that they were active only in their non-aggregated form. The Abeta response was saturable and not reduced by a substance P antagonist. This pharmacology excludes the participation of known Abeta binding proteins. The results indicate that a PI coupled receptor for non-aggregated Abeta may be present in brain.

Presenilin-1 expression in Pick's disease.

Recent studies have reported that neuronal populations expressing low levels of presenilin-1 (PS-1) display increased vulnerability in late-onset sporadic Alzheimer's disease (AD). To examine whether this phenomenon also occurs in other neurodegenerative diseases, we performed a quantitative immunocytochemical study of PS-1 distribution in the cerebral cortex of Pick's disease (PiD) cases and non-demented individuals. In PiD cases, the percentage of PS-1-containing, Pick body (PB)-free neurons was significantly elevated only in cortical areas showing neuronal loss. In these areas, PS-1 levels, measured by immunoblotting, were often higher in PiD compared to non-demented cases. Moreover, PS-1 immunoreactivity was significantly reduced in PB-containing neurons. These data suggest that as previously shown in AD, low cellular expression of PS-1 may be associated with increased neuronal loss and cellular degeneration.

Potentiation of the D2 mutant motor phenotype in mice lacking dopamine D2 and D3 receptors.

Within the D2-class of dopamine receptors, the D2 and D3 subtypes share the highest degree of similarity in their primary structure. However, the extent to which these two receptor subtypes have similar or different functional properties is unclear. The present study used gene targeting to generate mice deficient for D2, D3, and D2/D3 receptors. A comparative analysis of D2 and D3 single mutants and D2/D3 double mutants revealed that D2/D3 double mutants develop motor phenotypes that, although qualitatively similar to those seen in D2 single mutants, are significantly more severe. Furthermore, increased levels of the dopamine metabolites dihydroxyphenyl acetic acid and homovanillic acid are found in the dorsal striatum of D2 single mutants. The levels of these metabolites, however, are significantly higher in mice lacking D2 and D3 receptors. In addition, results of immunoprecipitation experiments revealed that D2 single mutants express higher levels of D3 receptor proteins during later stages of their postnatal development. These results suggest that D3 receptors compensate for some of the lacking D2 receptor functions and that these functional properties of D3 receptors, detected in mice with a D2 mutant genetic background, remain masked when the abundant D2 receptor is expressed.

Distinct secretases, a cysteine protease and a serine protease, generate the C termini of amyloid beta-proteins Abeta1-40 and Abeta1-42, respectively.

The carboxy-terminal ends of the 40- and 42-amino acids amyloid beta-protein (Abeta) may be generated by the action of at least two different proteases termed gamma(40)- and gamma(42)-secretase, respectively. To examine the cleavage specificity of the two proteases, we treated amyloid precursor protein (APP)-transfected cell cultures with several dipeptidyl aldehydes including N-benzyloxycarbonyl-Leu-leucinal (Z-LL-CHO) and the newly synthesized N-benzyloxycarbonyl-Val-leucinal (Z-VL-CHO). All dipeptidyl aldehydes tested inhibited production of both Abeta1-40 and Abeta1-42. Changes in the P1 and P2 residues of these aldehydes, however, indicated that the amino acids occupying these positions are important for the efficient inhibition of gamma-secretases. Peptidyl aldehydes inhibit both cysteine and serine proteases, suggesting that the two gamma-secretases belong to one of these mechanistic classes. To differentiate between the two classes of proteases, we treated our cultures with the specific cysteine protease inhibitor E-64d. This agent inhibited production of secreted Abeta1-40, with a concomitant accumulation of its cellular precursor indicating that gamma(40)-secretase is a cysteine protease. In contrast, this treatment increased production of secreted Abeta1-42. No inhibition of Abeta production was observed with the potent calpain inhibitor I (acetyl-Leu-Leu-norleucinal), suggesting that calpain is not involved. Together, these results indicate that gamma(40)-secretase is a cysteine protease distinct from calpain, whereas gamma(42)-secretase may be a serine protease. In addition, the two secretases may compete for the same substrate. Dipeptidyl aldehyde treatment of cultures transfected with APP carrying the Swedish mutation resulted in the accumulation of the beta-secretase C-terminal APP fragment and a decrease of the alpha-secretase C-terminal APP fragment, indicating that this mutation shifts APP cleavage from the alpha-secretase site to the beta-secretase site.

The kunitz protease inhibitor form of the amyloid precursor protein (KPI/APP) inhibits the proneuropeptide processing enzyme prohormone thiol protease (PTP). Colocalization of KPI/APP and PTP in secretory vesicles.

Proteolytic processing of proenkephalin and proneuropeptides is required for the production of active neurotransmitters and peptide hormones. Variations in the extent of proenkephalin processing in vivo suggest involvement of endogenous protease inhibitors. This study demonstrates that "protease nexin 2 (PN2)," the secreted form of the kunitz protease inhibitor (KPI) of the amyloid precursor protein (APP), potently inhibited the proenkephalin processing enzyme known as prohormone thiol protease (PTP), with a Ki,app of 400 nM. Moreover, PTP and PN2 formed SDS-stable complexes that are typical of kunitz protease inhibitor interactions with target proteases. In vivo, KPI/APP (120 kDa), as well as a truncated form of KPI/APP that resembles PN2 in apparent molecular mass (110 kDa), were colocalized with PTP and (Met)enkephalin in secretory vesicles of adrenal medulla (chromaffin granules). KPI/APP (110-120 kDa) was also detected in pituitary secretory vesicles that contain PTP. In chromaffin cells, calcium-dependent secretion of KPI/APP with PTP and (Met)enkephalin demonstrated the colocalization of these components in functional secretory vesicles. These results suggest a role for KPI/APP inhibition of PTP in regulated secretory vesicles. In addition, these results are the first to identify an endogenous protease target of KPI/APP, which is developmentally regulated in aging and Alzheimer's disease.

Presenilin-1 forms complexes with the cadherin/catenin cell-cell adhesion system and is recruited to intercellular and synaptic contacts.

In MDCK cells, presenilin-1 (PS1) accumulates at intercellular contacts where it colocalizes with components of the cadherin-based adherens junctions. PS1 fragments form complexes with E-cadherin, beta-catenin, and alpha-catenin, all components of adherens junctions. In confluent MDCK cells, PS1 forms complexes with cell surface E-cadherin; disruption of Ca(2+)-dependent cell-cell contacts reduces surface PS1 and the levels of PS1-E-cadherin complexes. PS1 overexpression in human kidney cells enhances cell-cell adhesion. Together, these data show that PS1 incorporates into the cadherin/catenin adhesion system and regulates cell-cell adhesion. PS1 concentrates at intercellular contacts in epithelial tissue; in brain, it forms complexes with both E- and N-cadherin and concentrates at synaptic adhesions. That PS1 is a constituent of the cadherin/catenin complex makes that complex a potential target for PS1 FAD mutations.

Enrichment of presenilin 1 peptides in neuronal large dense-core and somatodendritic clathrin-coated vesicles.

Presenilin 1 is an integral membrane protein specifically cleaved to yield an N-terminal and a C-terminal fragment, both membrane-associated. More than 40 presenilin 1 mutations have been linked to early-onset familial Alzheimer disease, although the mechanism by which these mutations induce the Alzheimer disease neuropathology is not clear. Presenilin 1 is expressed predominantly in neurons, suggesting that the familial Alzheimer disease mutants may compromise or change the neuronal function (s) of the wild-type protein. To elucidate the function of this protein, we studied its expression in neuronal vesicular systems using as models the chromaffin granules of the neuroendocrine chromaffin cells and the major categories of brain neuronal vesicles, including the small clear-core synaptic vesicles, the large dense-core vesicles, and the somatodendritic and nerve terminal clathrin-coated vesicles. Both the N- and C-terminal presenilin 1 proteolytic fragments were greatly enriched in chromaffin granule and neuronal large dense-core vesicle membranes, indicating that these fragments are targeted to these vesicles and may regulate the large dense-core vesicle-mediated secretion of neuropeptides and neurotransmitters at synaptic sites. The presenilin 1 fragments were also enriched in the somatodendritic clathrin-coated vesicle membranes, suggesting that they are targeted to the somatodendritic membrane, where they may regulate constitutive secretion and endocytosis. In contrast, these fragments were not enriched in the small clear-core synaptic vesicle or in the nerve terminal clathrin-coated vesicle membranes. Taken together, our data indicate that presenilin 1 proteolytic fragments are targeted to specific populations of neuronal vesicles where they may regulate vesicular function. Although full-length presenilin 1 was present in crude homogenates, it was not detected in any of the vesicles studied, indicating that, unlike the presenilin fragments, full-length protein may not have a vesicular function.

Overexpression in neurons of human presenilin-1 or a presenilin-1 familial Alzheimer disease mutant does not enhance apoptosis.

Programmed cell death, or apoptosis, has been implicated in Alzheimer's disease (AD). DNA damage was assessed in primary cortical neurons infected with herpes simplex virus (HSV) vectors expressing the familial Alzheimer's disease (FAD) gene presenilin-1 (PS-1) or an FAD mutant of this gene, A246E. After infection, immunoreactivity for PS-1 was shown to be enhanced in infected cells. The infected cells exhibited no cytotoxicity, as evaluated by trypan blue exclusion and mitochondrial function assays. Quantitative analysis of cells that were immunohistochemically labeled using a Klenow DNA fragmentation assay or the TUNEL method revealed no enhancement of apoptosis in PS-1-infected cells. This result was confirmed using assays for chromatin condensation and for DNA fragmentation. Expression of PS-1 protected against induction of apoptosis in the cortical neurons by etoposide or staurosporine. The specificity of this phenotype was demonstrated by the fact that cortical cultures infected with recombinant HSV vectors expressing the amyloid precursor protein (APP-695) showed, in contrast, a significant increase in the number of apoptotic cells and an increase in DNA fragmentation for all parameters tested. Our results indicate that overexpression of wild-type or A246E mutant PS-1 does not enhance apoptosis in postmitotic cortical cells and suggest that the previously reported enhancement of apoptosis by presenilins may be dependent on cell type.

Inhibition of Alzheimer beta-fibrillogenesis by melatonin.

It is generally postulated that the amyloid beta protein (Abeta) plays a central role in the progressive neurodegeneration observed in Alzheimer's disease. Important pathologic properties of this protein, such as neurotoxicity and resistance to proteolytic degradation, depend on the ability of Abeta to form beta-sheet structures or amyloid fibrils. We report that melatonin, a hormone recently found to protect neurons against Abeta toxicity, interacts with Abeta1-40 and Abeta1-42 and inhibits the progressive formation of beta-sheets and amyloid fibrils. These interactions between melatonin and the amyloid peptides were demonstrated by circular dichroism and electron microscopy for Abeta1-40 and Abeta1-42 and by nuclear magnetic resonance spectroscopy for Abeta1-40. Inhibition of beta-sheets and fibrils could not be accomplished in control experiments when a free radical scavenger or a melatonin analog were substituted for melatonin under otherwise identical conditions. In sharp contrast with conventional anti-oxidants and available anti-amyloidogenic compounds, melatonin crosses the blood-brain barrier, is relatively devoid of toxicity, and constitutes a potential new therapeutic agent in Alzheimer's disease.

Release of nontransmembrane full-length Alzheimer's amyloid precursor protein from the lumenar surface of chromaffin granule membranes.

We previously demonstrated the presence of a soluble form of full-length Alzheimer's amyloid precursor protein (APP) in the lumen of adrenal medullary chromaffin granules (CG). Furthermore, full-length APP is released from CG membranes in vitro at pH 9.0 by an enzymatic mechanism, sensitive to protease inhibitors [Vassilacopoulou et al. (1995) J. Neurochem. 64, 2140-2146]. In this study, we found that when intact CG were subjected to exogenous trypsin, a fraction of APP was not digested, consistent with an intragranular population of APP. To examine the substrate-product relationship between membrane and soluble full-length APP, we labeled CG transmembrane APP with 3-(trifluoromethyl)-3-(m-[125I]iodophenyl)diazirine ([125I]TID), a lipophilic probe, specific for membrane-spanning domains of proteins. APP released from the membranes at pH 9.0 was not labeled with [125I]TID. In addition, this APP was not biotinylated in intact CG. Combined, the results indicate that APP released from CG membranes derives from a unique nontransmembrane population of membrane-associated APP, located in the lumenal side of CG membranes. Dithiobis(succinimidylpropionate) (DSP) cross-linking indicated that APP in CG is situated in close proximity with other proteins, possibly with APP itself. APP complexes were also detected under nonreducing conditions, without DSP cross-linking. These results, combined with our previous studies, indicate that full-length APP within CG exists as three different populations: (I) transmembrane, (II) membrane-associated/nontransmembrane, and (III) soluble. The existence of nontransmembrane populations suggests that putative gamma-secretase cleavage sites of APP, assumed to be buried within the lipid bilayer, could be accessible to proteolysis in a soluble intravesicular environment.

Alzheimer's disease: a re-examination of the amyloid hypothesis.

Alzheimer's disease (AD) is a neurodegenerative disorder of the brain characterized by the presence of neuritic amyloid plaques and neurofibrillary tangles. Although it most frequently occurs in the elderly, this disorder also afflicts younger patients. The majority of AD cases are late in onset, lack an obvious genetic etiology and are characterized as sporadic, whereas a small percentage of cases are early in onset and segregate strongly within families (FAD), suggesting a genetic etiology. During the past decade it has become evident that the clinical and histopathological phenotypes of this disease are caused by heterogeneous genetic, and probably environmental, factors. Indeed, several genes have been identified that together appear to cause most of the familial forms of the disease, whereas the epsilon4 allele of the apolipoprotein E (apoE) gene has been shown to be a significant risk factor for the late onset forms of AD. Despite this evidence of heterogeneity, it has been suggested that all of these factors work through a common pathway by triggering the deposition of amyloid in the brain, which is ultimately responsible for the neuronal degeneration of AD. This is a controversial theory, however, primarily because there is a poor correlation between the concentrations and distribution of amyloid depositions in the brain and several parameters of AD pathology, including degree of dementia, loss of synapses, loss of neurons and abnormalities of the cytoskeleton.

Appican expression induces morphological changes in C6 glioma cells and promotes adhesion of neural cells to the extracellular matrix.

Appicans are secreted or cell-associated brain chondroitin sulfate proteoglycans produced by glia cells and containing Alzheimer amyloid precursor protein (APP) as a core protein. Here, we report that rat C6 glioma cells transfected with appican displayed a dramatic change in their phenotypic appearance compared with untransfected cells or cells transfected with APP. Appican-transfected cells lost the round appearance of the untransfected control C6 cells, acquired a flat morphology, and elaborated more processes than control cells. Untransfected, or APP-transfected C6, cells were completely dissociated from their substrate after 40 min of treatment with cell dissociation solution. Under the same conditions, however, <20% of the appican-transfected C6 cells were dissociated from their substrate, suggesting that the appican-transfected glia cells attach more avidly to their substrate than do untransfected or APP transfected control cells. In contrast, appican-transfected fibroblast cells showed no morphological changes and dissociated from their substrate similarly to untransfected fibroblast cells. Extracellular matrix (ECM) prepared from appican-transfected C6 cell cultures contained high levels of appican and was a significantly better substrate for the attachment of C6 cells than ECM from either untransfected or APP-transfected cultures. Furthermore, cell adhesion to ECM was independent of the level of appican expression of the plated cells. ECM from appican-transfected C6 cultures stimulated adhesion of other neural cells including primary astrocytes, Neuro2a neuroblastoma, and PC12 pheochromocytoma, but not fibroblast cells. Conditioned media from appican-transfected C6 cultures failed to promote cell adhesion. Together, these data suggest that secreted appican incorporates into ECM and promotes adhesion of neural cells. Furthermore, our data suggest that the chondroitin sulfate chain engenders APP with novel biological functions.

Melatonin prevents death of neuroblastoma cells exposed to the Alzheimer amyloid peptide.

Studies from several laboratories have generated evidence suggesting that oxidative stress is involved in the pathogenesis of Alzheimer's disease (AD). The finding that the amyloid beta protein (Abeta) has neurotoxic properties and that such effects are, in part, mediated by free radicals has provided insights into mechanisms of cell death in AD and an avenue to explore new therapeutic approaches. In this study we demonstrate that melatonin, a pineal hormone with recently established antioxidant properties, is remarkably effective in preventing death of cultured neuroblastoma cells as well as oxidative damage and intracellular Ca2+ increases induced by a cytotoxic fragment of Abeta. The effects of melatonin were extremely reproducible and corroborated by multiple quantitative methods, including cell viability studies by confocal laser microscopy, electron microscopy, and measurements of intracellular calcium levels. The importance of this finding is that, in contrast to conventional antioxidants, melatonin has a proposed physiological role in the aging process. Secretion levels of this hormone are decreased in aging and more severely reduced in AD. The reported phenomenon may be of therapeutic relevance in AD.

Presenilin-1-immunoreactive neurons are preserved in late-onset Alzheimer's disease.

Recent studies have suggested that missense mutations in the presenilin-1 gene are causally related to the majority of familial early-onset Alzheimer's disease (AD). To examine the possible involvement of presenilin-1 in late-onset sporadic AD, a quantitative analysis of its distribution in the cerebral cortex of nondemented and AD patients was performed using immunocytochemistry. Stereological analyses revealed that AD brains showed a marked neuronal loss in the CA1 field of the hippocampus and hilus of the dentate gyrus, subiculum, and entorhinal cortex. In these areas, however, the fraction of neurofibrillary tangle (NFT)-free neurons showing presenilin-1 immunoreactivity was increased compared with nondemented controls. In contrast, cortical areas, which displayed no neuronal loss, did not show any significant increase in the fraction of presenilin-1-positive neurons. Moreover, presenilin-1 immunoreactivity was reduced in NFT-containing neurons. Thus, in AD, the fraction of NFT-free neurons that contained presenilin-1 varied from 0.48 to 0.77, whereas the fraction of NFT-containing neurons that were presenilin-1 positive varied from 0.1 to 0.24. Together, these observations indicate that presenilin-1 may have a neuroprotective role and that in AD low cellular expression of this protein may be associated with increased neuronal loss and NFT formation.