| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
FJ
EXPRESS SUMMARY ARTICLE The Full-length version of this article is also available, published online September 13, 2004 as doi:10.1096/fj.03-1070fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1
* Department of Vascular Dementia Research, National Institute for Longevity Sciences, NCGG, Obu, Japan;
Department of Demyelinating Disease and Aging, National Institute of Neuroscience, NCNP, Kodaira, Japan; and
Tokyo Institute of Psychiatry, Tokyo, Japan
1Correspondence: E-mail: araki@ncnp.go.jp; tabira{at}nils.go.jp
SPECIFIC AIMS
Caspase activation and apoptosis are implicated in neuronal death in Alzheimer’s disease (AD). We analyzed the effects of the caspase-mediated cleavage of amyloid precursor protein (APP) on amyloid ß-peptide (Aß) production with special consideration of the generation of amino-terminally truncated Aß.
PRINCIPAL FINDINGS
1. Evidence for altered Aß generation from cells expressing caspase-cleaved form of APP
APP is cleaved by caspases in its cytoplasmic domain, subsequently generating APP lacking C-terminal 31 amino acids (APP
C). Human neuroblastoma SH-SY5Y cells stably transfected with wild-type APP or APPC were established and designated SH-APP and SH-APPC cells, respectively. We selected two pairs of SH-APP and SH-APPC cells expressing similar APP levels (designated SH-APP-1, -2, and SH-APPC-1, -2 cells). SH-APP-2 and SH-APPC-2 expressed 
twice as much APP as did SH-APP-1 and SH-APPC-1. Two types of sandwich ELISA were used to measure Aß in conditioned media from these cells. BNT77-based ELISA detected N-terminal-intact and truncated Aß (Aß40total and Aß42total), but not Aß17-40 (p3 fragment); BAN50-based ELISA detected only N-terminal-intact Aß (mainly Aß1-40 and Aß1-42). BNT77-based ELISA showed that SH-APP and SH-APPC cells secreted comparable levels of Aß40total and Aß42total. BAN50-based ELISA revealed that the amounts of Aß1-40 and Aß1-42 in SH-APPC cells were decreased to 30% of those found in SH-APP cells. These data suggest that N-terminally truncated Aß is increased relative to total Aß in SH-APPC cells.
We next analyzed C-terminal fragments (CTF) of APP by immunoprecipitation Western blot with 4G8 antibody. Two CTF bands (10 kDa 
-CTF and 12 kDa ß-CTF) and a faint band (ß'-CTF, 11 kDa) were detected in cell lysates of SH-APP. Similarly, two bands (6 kDa -CTFC and 8 kDa ß-CTFC) and a faint band (ß'-CTFC, 7 kDa) were observed in SH-APPC cell samples. The relative level of ß'-CTFC was increased in SH-APPC cells. Steady-state levels of ß-CTFC and -CTFC in SH-APPC cells were lower than those of ß-CTF and -CTF in SH-APP cells.
We compared the generation of secreted APP (sAPP) in SH-APP and SH-APPC cells. Immunoprecipitation Western blot showed that levels of total sAPP and sAPP- (sAPP derived from -secretase cleavage) were 4-fold higher in SH-APPC cells, compared with SH-APP cells.
2. Increased production of amino-truncated Aß from caspase-cleaved APP
We then analyzed altered Aß secretion in SH-APPC cells using immunoprecipitation Western blot. BAN50 antibody immunoprecipitated Aß1-40 and Aß1-42 (band 1, comigrating with synthetic Aß1-40/42). These immunoreactivities were decreased in media from SH-APPC cells, compared with SH-APP cell media. N-terminal-intact Aß (Aß1-40 and Aß1-42) and the smaller fragment (band 3, most likely Aß11-40 and Aß11-42) were detected in media from SH-APP cells by BNT77 immunoprecipitation. In contrast, the intensities of bands 1 and 3 were reduced, and the intensity of band 2 was increased in SH-APPC media (Fig. 1
A). We did not observe any band comigrating with Aß17-40 (p3 fragment) in the BNT77 immunoprecipitates.
|
To identify secreted Aß species, BNT77 immunoprecipitates were analyzed using a matrix-assisted laser desorption/ionization time-of-flight mass spectrometer (MALDI-TOF-MS). Two major peaks of Aß1-40 and Aß11-40 were detected in conditioned media from SH-APP cells. In contrast, the relative peak intensity of Aß5-40 was markedly increased, whereas peak intensities of Aß1-40 and Aß11-40 were decreased in SH-APPC media (Fig. 1B
). These data show that N-terminally truncated Aß (starting at Arg5) is markedly increased in media from SH-APPC cells. Aß11-40/42 appears to be generated through processing of APP by BACE1 and 
-secretase, as BACE1 alternatively cleaves between Tyr10-Glu11 in the Aß sequence. Increased levels of Aß5-40 and decreased levels of Aß1-40 and Aß11-40 were observed in media from HEK293 cells transiently transfected with APPC, compared with those transfected with APP.
We further analyzed the levels and species of intracellular Aß in SH-APP and SH-APPC cells by sensitive Western blot. Aß1-40 and Aß1-42 levels were comparable between samples from SH-APP and SH-APPC cells. The intracellular Aß1-42/Aß1-40 ratio was 0.4 in both cell lines. These results suggest that the generation of intracellular Aß is unaffected by the C-terminal truncation of APP.
3. Aß5-40/42 generation involves altered ß cleavage of APP
To determine the processing mechanism by which Aß5-40/42 is generated from APPC, we treated SH-APPC and SH-APP cells with a specific inhibitor of BACE, OM99-2. Secreted Aß was analyzed by immunoprecipitation Western blot and mass spectrometry. In OM99-2-treated SH-APPC cells, Aß1-40 secretion was significantly decreased, but the Aß5-40 level was not altered compared with that in untreated cells. SH-APP cells treated with the inhibitor secreted significantly reduced Aß1-40 and increased Aß5-40 levels. These data suggest that cleavage between Phe4 and Arg5 is not mediated by BACE1. BACE1 inhibition promotes the secretion of Aß5-40. To establish whether -secretase-like proteases are responsible for Aß5-40/42 generation, we incubated SH-APPC cells with a TACE (tumor necrosis factor- converting enzyme) inhibitor, TAPI-1, which inhibits -secretase. Incubation with 20 µM TAPI-1 resulted in increased Aß1-40 and decreased Aß5-40 secretion, suggesting that -secretase-like proteases are involved in Aß5-40/42 production.
4. Immunohistochemical analysis of Aß5-40/42 in AD brain
To determine whether Aß5-40/42 is present in human brain tissues, we generated a specific antibody to the N-terminal end region of this Aß species (designated the Aß5 antibody). In Western blot analyses, the Aß5 antibody reacted with Aß5-40, but not with Aß1-40, whereas the BAN50 antibody recognized only Aß1-40. The Aß5 antibody immunostained vessels in the AD brain, indicating the deposition of Aß5-40/42, particularly in vascular lesions with amyloid angiopathy. In nearby sections, another Aß antibody (6E10) labeled more vessels than did the Aß5 antibody. Amyloid angiopathy in the smaller sized vessels tended to be negative or weakly positive for Aß5. In addition, Aß5 antibody stained numerous neurofibrillary tangles (NFT), suggesting that Aß5-40/42 may be deposited in the NFT. Although a small number of senile plaques were positive for Aß5 in some cases, the staining was not as consistent as that of vessels and NFT.
5. Cleavage at the Aß5 site occurs when wild-type APP-expressing cells undergo apoptosis
Finally, we investigated whether APP processing to generate Aß5-40/42 occurs during apoptosis of wild-type APP-expressing cells. SH-APP cells were exposed to MG132 or MG132 plus staurosporine. Western blots using an AB5942 antibody specific for the caspase-generated neo epitope of APP reveal that exposure to these agents leads to caspase-mediated cleavage at the cytoplasmic region of APP. We examined CTF production from caspase-processed APP by immunoprecipitation Western blot analysis with 4G8 and AB5942 antibodies. Cells treated with MG132 plus staurosporine contained significant amounts of ß-CTFC, ß'-CTFC, and -CTFC, consistent with data from SH-APPC cells. The results suggest that cleavage at the Aß5 site occurs during apoptosis in SH-APP cells.
CONCLUSIONS AND SIGNIFICANCE
Recent evidence suggests that apoptosis underlies the neuronal death seen in AD. Active forms of caspases and the caspase-cleaved APP have been detected in AD brain tissues, but it is not clear whether the caspase cleavage of APP affects Aß formation. In this study, we have clearly demonstrated that such APP cleavage promotes the secretion of a distinct amino-truncated Aß species (Aß5-40/42). Our data provide the first evidentiary connection between caspase activation and the formation of amino-truncated Aß. Our results are consistent with and expand upon data from previous studies that measured N-terminally intact Aß, but not amino-truncated Aß.
We used inhibitors of BACE and -secretase to investigate the mechanism of Aß5-40/42 generation. After treatment of SH-APP cells with a BACE inhibitor, OM99-2, Aß1-40 levels were decreased, whereas Aß5-40 levels were increased. Treatment of SH-APPC cells with TAPI-1 led to decreased Aß5-40 and increased Aß1-40 levels. The data strongly suggest that cleavage at the Aß5 site is not ascribed to BACE1 activity, but mediated by -secretase-like proteases (e.g., ADAM family proteases, including TACE and ADAM10). This is consistent with the finding that secretion of p3 (Aß17-40), a product derived from -secretase cleavage, is significantly increased in cells expressing APPC. BACE2 functions as an alternative -secretase, but may not be involved in Aß5-40/42 generation, since OM99-2 inhibits BACE1 and BACE2.
It has been established that wild-type APP undergoes caspase cleavage during apoptosis, so it is reasonable to assume that subsequent cleavage at the Aß5 site occurs in apoptotic cells. We show evidence that APPC is generated in SH-APP cells exposed to MG-132 and staurosporine. ß'-CTFC, which corresponds to a precursor of Aß5-40/42, is formed in these apoptotic cells. Accordingly, we conclude that Aß5-40/42 is generated after caspase activation (Fig. 2
).
|
Our sensitive Western blot analyses indicate that the majority of intracellular Aß consist of Aß1-40/42 in wild-type APP- and APPC-expressing cells. Since two distinct pathways appear to exist for extracellular and intracellular pools of Aß, amino-truncated Aß peptides, including Aß5-40/42, are likely to be produced mainly in the extracellular Aß pathway.
Our immunohistochemical staining with the Aß5 antibody revealed that Aß5-40/42 species are deposited in some vessels with amyloid angiopathy in AD brain tissue. This may reflect the in vivo occurrence of caspase cleavage of APP. The observation that Aß5 antibody labels NFT is intriguing, considering that activation of caspases is suggested to occur in neurons bearing NFT. Our data are consistent with previous reports that considerable N-terminal modifications of Aß are seen in AD cortices and leptomeninges. Such amino-truncated Aß species may be instrumental in the amyloidosis process.
We suggest that caspase activation in the AD brain results in the formation of APPC, leading to the increased production and deposition of N-terminally truncated Aß5-40/42. Further research on the in vivo generation of this Aß species is needed to clarify its pathological role in Aß deposition and neuronal death in AD.
FOOTNOTES
To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.03-1070fje;
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
