高島 明彦

A polyclonal antibody, M5, to the hydrophilic loop domain of human presenilin 1 (PS1) was prepared. Western blot and immunoprecipitation analyses showed that M5 specifically recognized the processed C-terminal fragment, but not the full-length PS1. Epitope mapping analysis revealed that the essential sequence for recognition of the C-terminal fragment by M5 is DPEAQRR (302-308). The recognition of the C-terminal fragment by M5 in a processing-dependent manner was further confirmed by competitive enzyme-linked immunosorbent assay using the synthetic peptide L281 (281-311), which contains the putative processing site and the preceding amino acids to the site. Although L281 contains the epitope sequence for M5, the maximum inhibition was only 14%. Immunocytochemistry using M5 combined with hL312, which recognizes both full-length PS1 and the C-terminal fragment, allowed us to distinguish the localization of the processed C-terminal fragment from that of full-length PS1. Confocal microscopy demonstrated that the full-length form of wild-type PS1 is preferentially located in the nuclear envelope, while the processed C-terminal fragment is mainly present in the endoplasmic reticulum (ER). However, PS1 with familial Alzheimer's disease-associated mutations could not translocate to the nuclear envelope, and both the full-length and processed mutants were co-localized in the ER. Copyright (C) 2000 Elsevier Science Ireland Ltd and the Japan Neuroscience Society.

Numerous mutations causing early-onset familial Alzheimer's disease have been identified in the presenilin-1 gene. Presenilin-1 protein is produced as a 47 kDa holoprotein and proteolytically processed to an N-terminal 28 kDa and a C-terminal 19 kDa fragments by unidentified presenilinase in mammalian cells. We have demonstrated that this proteolytic processing also occurs in yeast. We also show that degradation of C-terminal fragment of presenilin-1 is dependent of proteasomal function. This yeast system will be a good tool to identify presenilinase and to study the role of presenilin-1 in amyloid precursor protein processing. Copyright (C) 2000 Elsevier Science Ireland Ltd.

Full-length form of human presenilin 1 (PS1) is processed and an N- terminal fragment (28 KD) and C-terminal fragment (19 KD) are generated. To elucidate the possible role of presenilin mutations in Alzheimer's disease (AD), the authors analyze the effects of AD-linked mutations on PSi processing in cultured cells. Complementary DNAs encoding genes for human PS1 harboring twenty-nine missense mutations linked with familial Alzheimer's disease (FAD) were introduced into PC12 cells. Human PS1 exogenously expressed in the cells was detected by immunoblotting using a monoclonal antibody that recognized the N-terminal region of human PS1. The amounts of full-length form (48 KD) and N-terminal fragment (28 KD) of PS1 was quantified by densitometrical analysis. The ratio of the N-terminal fragment to total PSI was reduced by twenty-nine mutations. The specific effects on PS1 processing varied according to mutation. These results suggest that AD- linked missense mutations of PS1 are involved in neurodegeneration, via inhibition of PS1 processing.

Families bearing mutations in the presenilin 1 (PS1) gene develop early onset familial Alzheimer's disease (FAD). Further, some PS1 mutants enhance secretion of the longer form of amyloid β protein (Aβ42). We constructed cDNAs encoding human PS1 harboring 28 FAD-linked mutations, and examined the effects of the expressed PS1 mutants on Aβ42 secretion in β amyloid precursor producing COS-1 cells. All the mutants significantly enhanced the ratio of Aβ42 to total Aβ compared with wild-type PS1. However, the increase in Aβ42 ratio in cells with each PS1 mutation did not correlate with the reported age of onset of FAD caused by that mutation. These results suggest that increased Aβ42 secretion is important for the development of Alzheimer's disease (AD), but may not be the only factor contributing to the onset of AD.

Many cases of early-onset familial Alzheimer's disease (FAD) are caused by mutations in the presenilin 1 (PS1) and PS2 genes. PS1 protein is generated as a 47 kDa protein and is endoproteolytically cleaved into N- terminal 28 kDa and C-terminal 19 kDa fragments in vivo. To examine whether mutated PS1 protein is abnormally metabolized, we performed immunoblot analysis of lymphoblasts from familial Alzheimer's disease patients and controls. More full-length PS1 was apparently detected in samples from PS1 mutants than those from PS2 mutant and controls. This result suggests that impaired proteolysis of PS1 may be associated with the pathogenesis of FAD. Moreover, our simple test using lymphocytes from FAD patients might be useful from a diagnostic point of view.

Missense mutations in the human presenilin-1 (PS1) gene, which is found on chromosome 14, cause early-onset familial Alzheimer's disease (FAD). FAD-linked PS1 variants alter proteolytic processing of the amyloid precursor protein and cause an increase in vulnerability to apoptosis induced by various cell stresses. However, the mechanisms responsible for these phenomena are not clear. Here we report that mutations in PS1 affect the unfolded-protein response (UPR), which responds to the increased amount of unfolded proteins that accumulate in the endoplasmic reticulum (ER) under conditions that cause ER stress. PS1 mutations also lead to decreased expression of GRP78/Bip, a molecular chaperone, present in the ER, that can enable protein folding. Interestingly, GRP78 levels are reduced in the brains of Alzheimer's disease patients. The downregulation of UPR signalling by PS1 mutations is caused by disturbed function of IRE1, which is the proximal sensor of conditions in the ER lumen. Overexpression of GRP78 in neuroblastoma cells bearing PS1 mutants almost completely restores resistance to ER stress to the level of cells expressing wild-type PS1. These results show that mutations in PS1 may increase vulnerability to ER stress by altering the UPR signalling pathway.

Presenilin 1 (PS1) has been identified as a causative gene for most early-onset familial Alzheimer's disease. Biochemical studies revealed that PS1 exists predominantly as two processed fragments in cells and brain tissues. We prepared stably transfected cells expressing the wild-type and familial Alzheimer's disease-associated mutants of PS1 and investigated the enzyme that participates in the metabolism of PS1. After treatment of the cells with proteasome inhibitors, the full-length PS1 was significantly accumulated. The levels of N- and C-terminal fragments were also increased. The accumulation of PS1 with a deletion of exon 10, which is unable to be processed, on treatment of the transfected cells with lactacystin indicated that proteasome can degrade full-length PS1. A synthetic peptide that includes the processing region of PS1 was cleaved by 20S proteasome at the putative processing sites after Met288 and Glu299. Metabolic labeling experiments showed that the appearance of the N-terminal fragment was attenuated by the inhibitor. Finally, 28-kDa N- and 20-kDa C-terminal fragments were generated by purified PS1 in vitro. These data indicated that the proteasome pathway is involved in PS1 processing. These results demonstrate that the proteasome pathway plays dual roles in processing and degradation of PS1.

Neural BC1 RNA is distributed in neuronal dendrites as ribonucleoprotein particles (RNP). Our previous studies indicated the presence of Translin in BC1 RNPs, which is a translational repressor and links a subset of mRNAs to microtubules. In this study, we confirmed that Translin associates with BC1 RNP and we used immunocytochemical methods to examine the subcellular distribution of Translin in developing hippocampal cells in primary cultures. Translin was detected in both the nuclei and cytoplasm of neurons, whereas in glial cells it was localized in the nuclei. Consistent with the reported developmental time course of BC1 RNA expression and dendritic delivery the translocation of Translin to the neuronal dendrites appeared to correlate with neuronal development and differentiation events such as the onset of synaptogenesis in culture. These observations suggest that BC1 RNP or Translin itself may be relevant to the dendritic translation of mRNAs in response to transsynaptic activity.

Families bearing mutations in the presenilin-1 (PS1) gene develop Alzheimer's disease (AD). However, the mechanism through which PS1 causes AD is unclear. The co-immunoprecipitation with PS1 in transfected COS-7 cells indicates that PS1 directly interacts with endogenous β-catenin, and the interaction requires residues 322-450 of PS1 and 445-676 of β-catenin. Both proteins are co-localized in the endoplasmic reticulum. Over-expression of PS1 reduces the level of cytoplasmic β-catenin, and inhibits β-catenin-T cell factor-regulated transcription. These results indicate that PS1 plays a role as inhibitor of the β-catenin signal, which may be connected with the AD dysfunction. Copyright (C) 1998 Federation of European Biochemical Societies.

Families bearing mutations in the presenilin 1 (PS1) gene develop Alzheimer's disease. Previous studies have shown that the AIzheimer- associated mutations in PS1 increase production of amyloid β protein (Aβ1-42). We now show that PS1 also regulates phosphorylation of the microtubule-associated protein tau. PS1 directly binds tau and a tau kinase, glycogen synthase kinase 3β (GSK-3β). Deletion studies show that both tau and GSK-3β bind to the same region of PS1, residues 250-298, whereas the binding domain on tau is the microtubule-binding repeat region. The ability of PSI to bring tau and GSK-3β into close proximity suggests that PS1 may regulate the interaction of tau with GSK-3β. Mutations in PS1 that cause Alzheimer's disease increase the ability of PS1 to bind GSK-3β and, correspondingly, increase its tau-directed kinase activity. We propose that the increased association of GSK-3β with mutant PS1 leads to increased phosphorylation of tau.

According to the amyloid hypothesis for the pathogenesis of Alzheimer's disease (AD), amyloid β peptide (Aβ) directly affects neurons, leading to neurodegeneration and tau phosphorylation, followed by the production of paired helical filaments (PHF) in neurofibrillary tangles (NFT). To analyze the relationship between the phosphorylation sites of tau and the activation of kinases in response to Aβ, we treated cultured rat hippocampal neurons with a peptide fragment of Aβ, Aβ(25-35). Aβ(25-35) treatment activated tau protein kinase I/glycogen synthase kinase-3β (TPK I/GSK-3β) but not glycogen synthase kinase-3α (GSK-3α) or mitogen activated protein kinase (MAP kinase) in primary culture of hippocampal neurons. Using antibodies that recognize phosphorylated sites of tau, we showed that tau phosphorylation was enhanced in at least five sites (Ser199, Ser202, Ser 396, Ser404, and Ser413 numbered according to the human tau isoform containing 441 amino acid residues), to an extent that depended on the level of TPK I/GSK-3β. Treatment with TPK I/GSK-3β antisense oligonucleotide inhibited the enhancement of tau phosphorylation induced by Aβ(25-35) exposure. Thus, TPK I/GSK-3β activation by AB(25-35) may lead to extensive tau phosphorylation.

Mutations in the gene for presenilin 1 are causative for the majority of cases of early onset familial Alzheimer's disease. Yet, the physiological function of presenilin 1 and the pathological mechanisms of the mutations leading to Alzheimer's disease are still unknown. To analyse potential pathological effects of presenilin 1 over-expression, we have generated transgenic rats which express high levels of human presenilin 1 protein in the brain. The over-expression of presenilin 1 leads to saturation of its normal processing and to the appearance of full-length protein in the transgenic rat brain. The transgenic protein is expressed throughout the brain and is predominantly found in neuronal cells. Cultured primary cortical neurons derived from these transgenic rats are significantly more sensitive than non-transgenic controls to apoptosis induced by standard culture conditions and to apoptosis induced by trophic factor withdrawal. Furthermore, the observed apoptosis is directly correlated with the expression of the transgenic protein. The results further emphasize the role of presenilin 1 in apoptotic cell death in native neuronal cultures.

One of the histopathological markers in Alzheimer's disease is the accumulation of hyperphosphorylated tau in neurons called neurofibrillary tangles (NFT) composing paired helical filaments (PHF). Combined tau protein kinase II (TPK II), which consists of CDK5 and its activator (p23), and glycogen synthase kinase-3β (GSK-3β) phosphorylate tau to the PHF-form in vitro. To investigate tau phosphorylation by these kinases in intact cells, the phosphorylation sites were examined in detail using well-characterized phosphorylation-dependent anti-tau antibodies after overexpressing the kinases in COS-7 cells with a human tau isoform. The overexpression of tau in COS-7 cells showed extensive phosphorylation at Ser-202 and Ser-404. The p23 overexpression induced a mobility shift of tau, but most of the phosphorylation sites overlapped the endogenous phosphorylation sites. GSK- 3β transfection showed the phosphorylation at Set-199, Thr-231, Ser-396, and Ser-413. Triplicated transfection resulted in phosphorylation of tau at 8 observed sites (Ser-199, Ser-202, Thr-205, Thr-231, Ser-235, Ser-396, Ser- 404, and Ser-413).

Tau protein kinases (TPK) I and II were isolated as candidate enzymes responsible for the hyperphosphorylation observed in PHF-τ. Four phosphorylation sites of tau were identified for each kinase, accounting for most, but not all, of the major phosphorylation sites of PHF-τ. Immunostaining with anti-TPKI antibody indicated that this kinase is up-regulated in AD brain. Such up-regulation of TPKI and phosphorylation of tau were reproduced by treating cultured hippocampal cells with amyloid β (Aβ) protein. In addition, we found that TPKI can phosphorylate and inactivate pyruvate dehydrogenase (PDH), which is expected to result in depletion of acetyl-CoA, a key substrate of acetyl choline synthesis. Indeed, when septum cells were treated with Aβ, the level of acetyl choline decreased dramatically.

We have cloned the rat homologue of the presenilin-2 (PS-2) cDNA. PS-2 is responsible for chromosome 1-linked familial Alzheimer's disease. Sequence analysis predicted that the rat PS-2 encodes a 448 amino acid (aa) protein, and there was a very high degree of amino acid identity between rat and human PS-2 (95%). All the mutated codons in PS-2 and PS-1 in chromosome 1- or 14-linked familial Alzheimer's disease patients were conserved in rat PS-2. The expression of PS-2 was weaker than that of PS-1. The alternatively spliced short form of PS-2 mRNA, which was detected in human tissues was not detected in various rat tissues. During brain development, the expression level of both PS-2 and PS-1 increased but decreased in the adult. No remarkable change was observed in neural differentiation of PC12 cells.

Presenilin 1 (PS 1) shows missense mutations in most early-onset familial Alzheimer's disease (FAD). Transfection of cDNA for wild type PSI into rat pheochromocytoma PC12 cells generated a 47 kDa full-size PS 1 protein, which was processed into a 28 kDa N-terminal fragment and a 19 kDa C-terminal fragment. We prepared selected AIzheimer-associated mutations (Gly384Ala, Leu392Val, and Cys410Tyr) of PS 1, which localized after a possible cleavage site. By transient expression in PC12 cells and rat glioma cell line, C6, we examined their influence on the processing of PS 1. Cys410Tyr inhibited proteolytic processing of PS 1, while Gly384Ala and Leu392Val did not. Thus, the Alzheimer related mutations can be divided into two groups in terms of their effect on the proteolytic cleavage of PS 1.

A series of single alanine substituted analogs of amyloid β peptide (25-35) were tested for their ability to activate the phospholipases of cultured LA-N-2 cells. Substitution of alanine for the amino acids 29-34 prevented the activation of phospholipases A2 and D. In addition substitution of alanine at 28 prevented phospholipase D but not phospholipase A2 activation. All the alanine substitutions, except for positions 33 and 35, blunted phospholipase C activations. There were no activations by scrambled amyloid β peptide.

Using well-characterized monoclonal antibodies which recognize the N-terminus of presenilin 1 (PS1), we examined by immunohistochemistry brain tissues from patients with nonneurological conditions, Alzheimer's disease (AD) and cerebral infarction, as well as normal human liver tissues. The antibodies to PS1 did not reveal any positive staining in nonneurological conditions. In AD, the antibody showed positive staining of intraneuronal neurofibrillary tangles and neuropil threads. In cerebral infarcts, some macrophages were positively stained. In contrast to the C-terminal fragment of PS1, which has been claimed to be present in senile plaques, the N-terminal fragment binds to intracellular and intradendritic pathological structures and may play a role in tau phosphorylation in AD. © 1997 Academic Press.

We show here that amyloid β peptide1-42 (Aβ1-42) may play a key role in the pathogenesis of the cholinergic dysfunction seen in Alzheimer's disease (AD), in addition to its putative role in amyloid plaque formation. Aβ1-42 freshly solubilized in water (non-aged Aβ1-42), which was not neurotoxic without preaggregation, suppressed acetylcholine (ACh) synthesis in cholinergic neurons at very low concentrations (10-100 nM), although non-aged Aβ1-40 was ineffective. Non-aged Aβ1-42 impaired pyruvate dehydrogenase (PDH) activity by activating mitochondrial τ protein kinase I/glycogen synthase kinase-3β, as we have already shown in hippocampal neurons (Hoshi, M., Takashima, A., Noguchi, K., Murayama, M., Sato, M., Kondo, S., Saitoh, Y., Ishiguro, K., Hoshino, T., and Imahori, K. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 2719-2723). Neither choline acetyltransferase activity nor choline metabolism was affected. Therefore, the major cause of reduced ACh synthesis was considered to be an inadequate supply of acetyl-CoA owing to PDH impairment. Soluble Aβ1-42 increases specifically in AD brain (Kuo, Y.-M., Emmerling, M. R., Vigo-Pelfrey, C., Kasunic, T. C., Kirkpatrick, J. B., Murdoch, G. H., Ball, M. J., and Roher, A. E. (1996) J. Biol. Chem. 271, 4077-4081). This increase in soluble Aβ1- 42 may disturb cholinergic function, leading to the deterioration of memory and cognitive function that is characteristic of AD.

5-Azacytidine (5 Az) is a potent inhibitor of DNA methylation, and it may allow inactive genes to become expressed. In a previous study, we demonstrated that 5 Az administered to the dam induced apoptosis in the brains of fetal mice. In this study, the 5 Az-induced apoptosis was further characterized in differentiated PC 12 cells as a model for neuronal apoptosis. Cell death, determined by the activity of released lactate dehydrogenase (LDH) into the medium, occurred from 24 to 48 hrs after 5 Az treatment. Toxicity for differentiated PC 12 cells was observed on treatment with more than 10-1 μg/ml of 5 Az, and it reached the maximal level at 10-1 μg/ml. Cycloheximide, an inhibitor of protein synthesis, prevented 5 Az toxicity, suggesting that this cell death required protein synthesis which could be related to the activation of a dormant gene(s). Electrophoresis of DNA from 5 Az-treated cells evoked ladder formation, indicating the cleavage of DNA into nucleosomes. Scanning electron microscopy demonstrated bleb formation, the so-called apoptotic bodies on the cell surface. The biochemical and morphological findings indicated that 5 Az-induced cell death occurred in the form of apoptosis. 5 Az-induced cell death was prevented by treatment with cAMP but not by treatment with high K+ or deoxycytidine. These results suggest that a cAMP-sensitive mechanism is involved in 5 Az-induced cell death. PC 12 cells should be of value in elucidating the molecular mechanism of 5 Az-induced neuronal apoptosis.

The three-dimensional structure of amyloid β peptide (25-35), which has neurotoxic activity, in lithium dodecyl sulfate micelles was determined by two-dimensional 1H NMR spectroscopy with simulated annealing calculations. A total of 20 converged amyloid β peptide structures were obtained on the basis of 110 experimental constraints, including 106 distance constraints reduced from the nuclear. Overhauser effect (NOE) connectivities and four torsion angle (φ) constraints. The atomic root mean square difference about averaged coordinates is 1.04 ± 0.25 Å for the backbone atoms (N, C(α), C) and 1.39 ± 0.27 Å for all heavy atoms of the entire peptide. The molecular structure of amyloid β peptide in membrane- mimicking environment is composed of a short α helix in the C terminal position. The three residues from the N-terminus are disordered, but the remaining eight C-terminal residues are well-ordered, which is supported by the RMSD values of the C-terminal region, Lys28-Leu34. In this region, the RMS differences from averaged coordinates are 0.26 ± 0.11 Å for the backbone atoms (N, C(α), C) and 0.77 ± 0.21 Å for all heavy atoms, which is very low compared with those for the entire peptide. The four amino acid residues from the N-terminus are hydrophilic and the other seven amino acid residues in C-terminus are hydrophobic. So, our results show that the C- terminal region of amyloid β peptide (25-35) is buried in the membrane and assumes α-helical structure, whereas the N-terminal region is exposed to the solvent with a flexible structure. This structure is very similar to membrane-mediated structure of substance P previously reported. The three- dimensional structure of a non-neurotoxic mutant of amyloid β peptide (25- 35), where Asn27 is replaced by Ala, in lithium dodecyl sulfate micelles was also determined. The structure is similar to that of the wild type amyloid β peptide (25-35) in the C-terminal region, but the N-terminal flexible region is different. The structural comparison of amyloid β peptide (25-35), its non-neurotoxic mutant and substance P gives a structural basis to understand the mechanism of neurotoxicity caused by amyloid β peptide.

Presenilin 1 (PS 1) is the recently identified gene, located on chromosome 14, of which missense mutations can cause early-onset familial Alzheimer's disease. To understand the normal biological function of presenilin 1, we examined the sub-cellular localization by using a monoclonal anti-presenilin 1 antibody. Immuno-electronmicroscopic and biochemical analysis indicated that presenilin 1 is localized on cellular membrane (plasma, endoplasmic reticulum, and perinuclear) in COS-7 cells overexpressing presenilin 1. Interestingly, the PS 1 immunoreactivity in the plasma membrane was concentrated in the regions with cell-cell contact. This observation suggests a possible role of PS 1 on the cell membrane as a cell adhesion molecule.

Using immunohistochemistry, we examined the localization of four types of proline-directed kinases in the brains of control rats and in the brains of non-demented aged human subjects, subjects with Alzheimer's disease and those with Down's syndrome. The four kinases were: cyclin-dependent kinase (cdk) 5, a component of tau protein kinase (TPK) II TPK I/glycogen synthase kinase (GSK)-3β GSK-3α and mitogen-activated protein kinase (MAPK/ERK2). Each of these kinases has been reported to promote the hyperphosphorylation of tau protein in vitro. The kinases were located essentially in neurons, although the intensity and distribution of labeling varied. Antiserum for cdk5 showed the most preferential and consistent labeling of intraneuronal neurofibrillary tangles (NFT). Antiserum for TPK I/GSK-3β also labeled intraneuronal NFT. Double immunolabeling for TPK I/GSK-3β and tau1 showed that TPK I/GSK-3β was closely associated with NFT. Antiserum for GSK-3α labeled neurons weakly, and the intensity of labeling did not differ between neurons with and without NFT. Antiserum for MAPK labeled neurons in superficial cortical layers, but NFT appeared in both superficial and deep cortical layers. These findings suggest that cdk5 and TPK I/GSK-3β are the critically important kinases for the generation in vivo of hyperphosphorylated tau, the main component of the paired helical filaments in NFT.

The majority of cases of early-onset familial Alzheimer disease are caused by mutations in the recently identified presenilin 1 (PS1) gene, located on chromosome 14. PS1, a 467 amino acid protein, is predicted to be an integral membrane protein containing seven putative transmembrane domains and a large hydrophilic loop between the sixth and seventh membrane-spanning domain. We produced 7 monoclonal antibodies that react with 3 non-overlapping epitopes on the N-terminal hydrophilic tail of PS1. The monoclonal antibodies can detect the full-size PS1 at M(r) 47 000 and a more abundant M(r) 28 000 product in membrane extracts from human brain and human cell lines. PC12 cells transiently transfected with PS1 constructs containing two different Alzheimer mutations fail to generate the 28 kDa degradation product in contrast to PC12 cells transfected with wild-type PS1. Our results indicate that missense mutations in this form of familial Alzheimer disease may act via a mechanism of impaired proteolytic processing of PS1.

According to the amyloid hypothesis for the pathogenesis of Alzheimer disease, β-amyloid peptide (βA) directly affects neurons, leading to neurodegeneration and tau phosphorylation. In rat hippocampal culture, βA exposure activates tau protein kinase I/glycogen synthase kinase 3β (TPKI/GSK-3β), which phosphorylates tau protein into Alzheimer disease-like forms, resulting in neuronal death. To elucidate the mechanism of βA- induced neuronal death, we searched for substrates of TPKI/GSK-3β in a two- hybrid system and identified pyruvate dehydrogenase (PDH), which converts pyruvate to acetyl-CoA in mitochondria. PDH was phosphorylated and inactivated by TPKI/GSK-3β in vitro and also in βA-treated hippocampal cultures, resulting in mitochondrial dysfunction, which would contribute to neuronal death. In cholinergic neurons, βA impaired acetylcholine synthesis without affecting choline acetyltransferase activity, which suggests that PDH is inactivated by βA-induced TPKI/GSK-3β. Thus, TPKI/GSK-3β regulates PDH and participates in energy metabolism and acetylcholine synthesis. These results suggest that TPKI/GSK-3β plays a key role in the pathogenesis of Alzheimer disease.

The rat homologue of the presenilin-1 (PS-1) gene, which is responsible for early-onset familial Alzheimer's disease linked to chromosome 14, was cloned and sequenced. The predicted amino acid sequence showed quite high homology among rat, mouse, and human PS-1. Especially, the amino acid sequences of the putative transmembrane domains were highly conserved among the three species. The expression level of the PS-1 gene increased during brain development and the number of transcripts of the PS-1 gene changed during brain development. We found one transcript of the PS-1 gene in embryonic day 12 (E12)-E15 rat brain and two transcripts in E18-adult rat brain. Therefore, PS-1 may play a role in neurogenesis.

Exposure of rat hippocampal neurons to the peptide amyloid β (Aβ) (25-35) as well as Aβ (1-40) peptides enhances phosphorylation of tau to a paired helical filament (PHF)-state through activation of tau protein kinase I (TPK I)/glycogen synthase kinase-3β (GSK-3β) [Busciglio, J., Lorenzo, A., Yeh, J. and Yankner, B.A., Neuron, 14 (1995) 879-888 Takashima, A., Ishiguro, K., Noguchi, K., Michel, G., Hoshi, M., Sato, K., Takahashi, M., Hoshino, T., Uchida, T. and Imahori, K., Neurosci. Meeting Abstr., 671 (1995) 17]. In order to examine the effects of Aβ treatment on intracellular signaling mechanism, we have investigated the role of phosphatidyl inositol-3 (PI-3) kinase in the phosphorylation of tau. Aβ (25-35) exposure induced an inactivation of PI-3 kinase and an activation of TPK I/GSK-3β in rat hippocampal culture. Wortmannin, an inhibitor of PI-3 kinase, also activated TPK I/GSK-3β, leading to an enhancement of tau phosphorylation and neuronal death in hippocampal culture. These results suggest that Aβ (25-35) inhibition of PI-3 kinase results in the activation of TPK I/GSK-3β, the phosphorylation of tau, and resultant neuronal death in rat hippocampal neurons.

Exogenous application of synthetic amyloid β protein (Aβ) is known to induce neurotoxic effects in rat hippocampal culture. We report here that Aβ (25-35) induces accumulation of amyloid precursor protein (APP) derivatives in the cytoplasm of neurons. At the same time, the level of the secreted form of APP released into the culture medium decreases. Tau protein kinase I/glycogen synthase kinase-3β (TPK I/GSK-3β) antisense oligonucleotide blocked APP accumulation and prevented neuronal death. These results provide evidence that APP accumulation after Aβ treatment is regulated by TPK I/GSK-3β. Aβ neurotoxicity is probably mediated via phosphorylation of tau by TPK I/GSK-3β, resulting in an impairment of axonal transport, and cytoplasmic accumulation of APP. © 1995.

Structure-neurotoxicity relationships of amyloid β(25-35) peptide were studied by replacing each amino acid with Ala. In contrast to the general tendency in hydrophobicity-toxicity relationships, replacement of Asn27 yielded a more hydrophobic but less toxic analog and that of Met35 gave a less hydrophobic but more toxic one. Sedimentation profiles and CD spectra indicated that peptide aggregation via intermolecular β- sheet formation is essential for the neurotoxicity of amyloid β(25-35) peptide. The correlation between neurotoxicity and amyloid precursor protein accumulation suggested that the latter is one of the pathways of the neuronal death caused by amyloid β protein. © 1995 Oxford University Press.

We examined by immunohistochemicistry the carboxyl (C)-terminus extent of the amyloid beta protein (Aβ that constitutes senile plaques and amyloid angiopathy in the brains of non-demented and Alzheimer's disease (AD) subjects. We developed two antisera, which selectively recognized free C-termini of Aβ: BC40 for Aβ40 and BC42 for Aβ42. BC42 labeled various types of senile plaques as well as reference Aβ antiserum, whereas only some parts of the senile plaques were positive with BC40: i.e., all of the plaque cores and some diffuse and primitive plaques. In the brains of non-demented middle-aged subjects, a majority of BC42-positive diffuse plaques were also positive with BC40, but with less intensity than that shown with BC42. The ratio of BC40-negative plaques increased with increasing plaque density. Amyloid in the meningeal vessels showed much greater immnnoreactivity with BC40 than with BC42. Some extracellular neurofibrillary tangles were positive with both BC40 and BC42, although most of them and all the intraneuronal tangles showed no immunoreactivity with either antiserum. Aβ42 contributed exclusively to senile plaque formation, while contribution of Aβ40 varied among subjects. In vascular amyloid. Aβ40 is an essential component, while Aβ42 showed individual variation. © 1995 Informa UK Ltd All rights reserved: reproduction in whole or part not permitted.

We examined the subcellular distribution of two glycogen synthase kinase-3 (GSK-3) isoforms in rat cerebellum. Results from immunoelectron microscopy and subcellular fractionation revealed that one isoform, tau protein kinase I/GSK-3β (TPKI/GSK-3β), was present in mitochondria, but GSK-3α was not. Although the two GSK-3 isoforms seem to have similar properties, the difference of subcellular localization observed here suggests that TPKI/GSK-3β fulfills some specific function in mitochondria. © 1995 Oxford University Press.

We have studied the relationships of cell growth to muscarinic stimulation of norepinephrine release and phosphoinositide hydrolysis in the rat pheochromocytoma PC12 cells. The ability of these cells to release norepinephrine in response to muscarinic agonists was maximal during the early phase of exponential growth, and then rapidly decreased to undetectable levels as the cells approached stationary phase. In contrast, muscarinic stimulation of phosphoinositide hydrolysis was low in the early exponential phase of growth, increased to a maximum during late exponential growth and then dramatically dropped in the stationary phase. The number of muscarinic receptors, as measured by antagonist-binding studies, also varied during cell growth with maximal levels at days 2 and 8, corresponding to the maxima in muscarinic-stimulated norepinephrine release and phosphoinositide hydrolysis, respectively. © 1993.

Pathological changes of Alzheimer disease are characterized by cerebral cortical atrophy as a result of degeneration and loss of neurons. Typical histological lesions include numerous senile plaques composed of deposits of amyloid β-protein and neurofibrillary tangles consisting predominantly of ubiquitin and highly phosphorylated tau proteins. Previously, tau protein kinase I (TPK I) was purified and its cDNA was cloned. To examine the biological role of this enzyme in neurons, we have studied the induction of its kinase activity in primary cultures of embryonic rat hippocampal neurons. Treatment of cultures with amyloid β-protein significantly increased TPK I activity and induced the appearance of tau proteins recognized by the Alz-50 monoclonal antibody. In addition, though amyloid β-protein was neurotoxic, either cycloheximide or actinomycin D prevented neuronal death. Death was also prevented by TPK I antisense oligonucleotides but not by sense oligonucleotides. These observations suggest that rat hippocampal neurons undergo programmed cell death in response to amyloid β-protein and that TPK I is a key enzyme in this process.

The effects of the phorbol ester, 12-O-tetradecanoylphorbol 13-acetate (TPA) on stimulus-evoked dopamine release were studied in PC12 cells. Pretreatment of the cells with TPA resulted in an enhancement of dopamine release which could be further stimulated by high concentrations of K+, A23187, but not with carbamylcholine. TPA-dependent, high-K+-evoked enhancement of dopamine release was studied in detail: a maximum release was observed (169% of control) in response to 50 mM KCl upon treatment with 10-7 M TPA for 5 min at 37°C. This enhancement of dopamine release was associated with the concomitant reduction of the concentration rise of intracellular Ca2+ ([Ca2+]i) induced by a high concentration of K+ monitored by a fluorescent indicator, fura2. Thus, these data provide an example for alteration in the efficiency of stimulus-secretion coupling as pointed out in our previous paper. Moreover, we have shown that nicardipine, CdCl2, and CoCl2 inhibit high-K+-evoked dopamine release more effectively in TPA treated cells than that of untreated cells, and that the TPA-dependent, high-K+-evoked dopamine release observed in TPA treated cells is completely abolished by the presence of nicardipine, Cd2+ or Co2+, but is only partially inhibited in the presence of verapamil. These relevant findings suggest the possible involvement of protein kinase C in regulating the efficiency of a high-K+-evoked dopamine release through the modification of nicardipine-sensitive Ca2+ channels. © 1987.

Abstract: Catecholamine biosynthesis and its stimulusevoked release in PC12 pheochromocytoma cells were studied as a function of cell cycle by means of HPLC with electrochemical detection. We found that 3,4‐dihydroxyphenylethylamine (dopamine) levels in PC12 cells remained constant throughout the period of cell cycle. In contrast, the noradrenaline content was dependent on the cell cycle: it increased during the S + G2 phase followed by a decrease in the M phase. These results were confirmed further by measuring the activities catalyzing the catecholamine biosynthesis. Thus, activities of tyrosine 3‐monooxygenase and 3,4‐dihydroxyphenylalanine decarboxylase were independent of the cell cycle, whereas both soluble and membrane‐bound dopamine β‐monooxygenase activities were modulated during the cell cycle. On the other hand, release of the catecholamines stimulated with 50 mM KCI increased in the G1 phase, reached a maximum in the late G1, and then gradually decreased in later periods. We also found that carbamylcholine‐induced release of the catecholamines occurred maximally in the early S + G2 phase followed by a decrease during the M phase. Cell cycle dependence of the catecholamine release was in good agreement with that of 45Ca2+ uptake. Thus, this study provides evidence that the catecholamine biosynthesis and its release in PC12 cells are modulated during the period of cell cycle. Copyright © 1986, Wiley Blackwell. All rights reserved

The PC12 cell line derived from a rat adrenal medullary tumor is known to synthesize dopamine and to release it in response to cholinergic agonists or depolarizing agents. In this report, we have studied the relationship between dopamine biosynthesis and its stimulus-induced secretion in PC12 cells as a function of cell growth. The endogenous dopamine content was found to depend on cell growth, and reached a maximum in the stationary phase. This increase was associated both with an increase in the specific activity of tyrosine 3-monooxygenase, and with an increase of DOPA-decarboxylase in the cells. On the other hand, the maximal release of dopamine occurred in the late exponential phase before the endogenous dopamine was maximally synthesized in the cells. Moreover, the uptake of 45Ca2+ stimulated with either carbamylcholine or high K+ was also regulated by cell division: the maximal uptake took place in the same period of culture in which the maximal release of dopamine was observed. Thus, this report offers new evidence that the biosynthesis and secretion of dopamine are separately regulated in PC12 cells. © 1985.

Abstract: Clonal variants of PC12 cells with respect to catecholamine biosynthesis were isolated, and the catecholamine content was measured by high performance liquid chromatography with electrochemical detection. The dopamine content of 13 subclones, which were selected and isolated in tyrosine‐free medium, was substantially higher than the control level: 0.91 ± 0.10 nmol/mg protein (mean ± SEM n = 3). In contrast, the noradrenaline content showed a marked heterogeneity: only two subclones contained noradrenaline levels similar to or higher than the control level: 0.40 ± 0.05 (n = 5). The rest of them contained below the level of 0.20, and only negligible amounts of noradrenaline were found in four subclones. Thus, the noradrenaline‐to‐dopamine ratio varied widely between 0.003:1 and 0.53:1. This divergence of the noradrenaline content appears to be related to differing levels of dopamine β‐monooxygenase activity. The administration of ascorbate to the medium alone, however, did not restore the level of noradrenaline to the normal level in a subclone. Heterogeneity of the response to applied glucocorticoid was also demonstrated. Copyright © 1984, Wiley Blackwell. All rights reserved