University of Tennessee, Knoxville TRACE: Tennessee Research and Creative Exchange
Doctoral Dissertations Graduate School
5-2005
Identification of a novel cleavage site within the transmembrane domain of β-amyloid precursor protein and its implications in Aβ generation
Guojun Zhao University of Tennessee, Knoxville
Follow this and additional works at: https://trace.tennessee.edu/utk_graddiss
Recommended Citation Zhao, Guojun, "Identification of a novel cleavage site within the transmembrane domain of β-amyloid precursor protein and its implications in Aβ generation. " PhD diss., University of Tennessee, 2005. https://trace.tennessee.edu/utk_graddiss/4349
This Dissertation is brought to you for free and open access by the Graduate School at TRACE: Tennessee Research and Creative Exchange. It has been accepted for inclusion in Doctoral Dissertations by an authorized administrator of TRACE: Tennessee Research and Creative Exchange. For more information, please contact [email protected]. To the Graduate Council:
I am submitting herewith a dissertation written by Guojun Zhao entitled "Identification of a novel cleavage site within the transmembrane domain of β-amyloid precursor protein and its implications in Aβ generation." I have examined the final electronic copy of this dissertation for form and content and recommend that it be accepted in partial fulfillment of the equirr ements for the degree of Doctor of Philosophy, with a major in Comparative and Experimental Medicine.
Xuemin Xu, Major Professor
We have read this dissertation and recommend its acceptance:
Mei-Zhen Cui, Robert Donnell, Karla Mattesson, Hildegard M. Schuller
Accepted for the Council:
Carolyn R. Hodges
Vice Provost and Dean of the Graduate School
(Original signatures are on file with official studentecor r ds.) To the Graduate Council:
I am submitting herewith a dissertation written by Guojun Zhao entitled "Identification of a novel cleavage site within the transmembrane domain of �-amyloid precursor protein and its implications in A� generation". I have examined the final paper copy of this dissertation for formand content andrecommend that it be accepted in partialfulfillment of the requirements for the degree of Doctor of Philosophy, with a major in Comparative and Experimental Medicine.
Xuemin Xu, Major Professor
We have read this dissertation andrecommend its acceptance:
Acceptance forthe Council:
Graduate Studies - 2.c b IDENTIFICATION OF A NOVEL CLEAVAGE SITE WITHIN THE TRANSMEMBRANE DOMAIN OF P AMYLOID PRECURSOR PR,OTEIN AND ITS IMPLICATIONS IN Ap GENERATION
A Dissertation Presented for the Doctor of Philosophy Degree
The University of Tennessee, Knoxville
Guojun ZHAO
May 2005 DEDICATION
This dissertation is dedicated to:
My Wife Hui LI
and
My Son Tong ZHAO
11 ACKNOWLEDGEMENTS
I would like to thank my major professorDr. Xuemin Xu and co-supervisor Dr. Mei Zhen Cui for their overall directing and support during the course of my research. I would like to express my sincere appreciation to the other committee members, Dr. Hildegard M. Schuller, Dr. Karla Mattesson, and Dr. Robert Donnell for their valuable advice in both academic study and dissertation writing. I would also like to acknowledge Debbie Hampstead for valuable suggestions on graduate study and Dr. Stephen Kania for monitoring my comprehensive examination. I wish to thank all of my colleagues for their help both in technical aspects and in my daily life, especially during a time of personal illness. They are Y ongchang Shi, Wei Gao, Jeffrey Wiles, Guozhang Mao, Yunzhou Dong, Jianxin Tan, Longsheng Sun, Mingqin Tan, Feng Hao, EssamLaag, Keiko Eda, Ying Hung, and Ping Song. I am very grateful to my master degree supervisor Professor Y ongming Jiang and ProfessorDexin Sui for giving me precious training in biochemistry. Lastly, I own the greatest debts to my wife Hui Li and my lively son Tong Zhao for their consistent support and encouragement thatmade my dream come true.
iii ABSTRACT
Accumulation of A� is widely believed to play an early and crucial role in the pathogenesis of Alzheimer's disease (AD). The central problem in AD research, the mechanism of A� generation mediated by y-secretase, remains unclear. The present study has identified a novel cleavage site, s-cleavage, within the transmembrane domain of amyloid precursor protein (APP) and examined the role of s-cleavage in A� generation.
A novel A� species is produced without y-secretase inhibitor treatment, but is increased with the treatments in several cell lines. Both MALDI-Mass spectrum and western blot demonstrate that the novel A� is A�46. Both a dominant negative form of PS1 (D385A) and double deficiencies of PS1/PS2 abolish A�46 generation, demonstrating the dependence of s-cleavage on PS. The Indiana mutation of APP enhances A�46 generation. Interestingly, transition state analogues L685,458 and 31C block A�46 generation, whereas applications of non-transition state y-secretase inhibitors l�ad to an elevated level of A�46.
A�40/42 can be generated from pre-accumulated A�46 in the presence of L685,458.
A�46 is co-immunoprecipicated with PSl. Moreover, A� species which shows the same
migration rate in SDS-PAGE as synthetic A�49 was observed, suggesting that it is the long-sought A�49. Artificially-made A�49 is processed into A�46. Inhibition of A�46 generation from A�49 by L685,458 diminishes A�40/42 production. These findings suggest the sequence of e-, s-, and y-cleavages within the TM domain of APP.
A�46 is present in 20,000 x g pellet. The majority of A�46 is resistant to cell surface trypsinization. Treatment with BF A completely blocks A�46 production. A high dose of
IV monesin causes partial inhibition of AP46 generation, whereas treatments with NH4Cl, lactcysitn, or tunicamycin do not affectAP46 level. Furthermore, retention of APP in the
ER with an ER sorting signal abolishes AP46 generation, whereas addition of a TGN sorting motif does not affect AP46 level. These observations demonstrate that the major subcellularsites forAP46 generation are the Golgi and TGN.
Collectively, the present study provides new insights into the underlying mechanism of
AP generation.
V TABLE OF CONTENTS
Chapter 1 Molecular basis of Alzheimer's disease...... •.... 1
Chapter 2 Identification of a new presenilin-dependent ;-cleavage site within the transmem brane domain of amyloid precursor protein...... •...... •...... 22
Chapter 3 Amyloid precursor protein is processed within its transmembrane domain by sequential .:-, ;-, and y-cleavages...... 36
Chapter 4 Characterization of y-secretase inhibitors: their effects on Ap46 accumulation and turnover ...... •...... ••...... 55
Chapter 5 AP46 is generated in the golgi/trans-golgi network in neuro 2a cells ..... 72
References ...... •.•...... •...... • 93
Appendix...... •...... •...... 107
Vita •...... •...•...... •...... 136
VI LIST OF TABLE
Table 1 y-secretase inhibitor information...... 108
Vll LIST OF FIGURES Chapter 1 Fig.1. Some known mutations in APP ...... •...... •..... 109
Chapter 2 Fig.1. Detection of a new intracellular AP species ...... 110
Chapter 2 Fig.2. Detection of the novel AP species from several cell lines ...... 111
Chapter 2 Fig.3. MALDI-MS spectrum for the novel AP species ...... • 112
Chapter 2 Fig.4. Dominant negative form of PSl diminishes Ap46 generation ...... 11 3
Chapter 2 Fig.5. Double knockout of PS1/PS2 abolishes AP46 generation ...... 11 4
Chapter 2 Fig.6. Deferential inhibitions of AP46 formation by transition state analogs and non-transition state inhibitors ...... •...... •11 5
Chapter 2 Fig.7. The Indiana mutation of APP enhances Ap46 generation ...... 116
Chapter 3 Fig.1. The absence of AP46 in cells treated with L-685,458 is due to its
inhibition of the formation of A�46 ...•...... •...•..•...... 11 7
Chapter 3 Fig.2. AP46 turnover in a cell-freesystem ...... 118
Chapter 3 Fig.3. Generation of secreted AP40/42 from AJl46...... 119
Chapter 3 Fig.4. Schematic diagram of the procedure of crude membrane
preparation and solubilization ...... 120
Chapter 3 Fig.5. AP46 is associated with PSl...... •.....121
Chapter 3 Fig.6. Detection of AP49 in intact cells and the production of secreted
AP40/42 from AP49 ...... 122
Chapter 3 Fig.7. Schematic illustration of APP processing ...•...... •...... 123
Chapter 4 Fig.1. Effects of DAPT,compound E, ibuprofen, and sulindac sulfide on
AJ146 and CTFP accumulation ...... 124
Chapter 4 Fig.2. Inhibition profiles of various y-secretase inhibitors on APP
Vlll processing ...... 12 5
Chapter 4 Fig.3. Effects of inhibitor concentration on AP42/40 ratio ...... 126
Chapter 4 Fig.4. The profileof AP46 and CTFP after removal of y-secretase inhibitors...... 127
Chapter 4 Fig.5. L685,458 and 31C differentially affected AP46 turnover...... 128
Chapter 5 Fig.1. AP46 is present in 20kxgmembrane fraction ...... 129
Chapter 5 Fig.2. The majority of AP46 is resistent to surface trypsinization...... 130
Chapter 5 Fig.3. Treatments of BFA and monensin lead to distribution of APPswe into the ER and TGN in N2a cells, respectively ...... •...... 131
Chapter 5 Fig.4. The effects of BFA, monensin, lactasystin, NH4Cl and tunicamycin on soluble APP secretion ...... 13 2
Chapter 5 Fig.5. The effectsof BFA, monensin, lactacystin, NH4Cl, and tunicamycin on y- secretase mediated APP cleavages ...... •...... 133
Chapter 5 Fig.6. Addition of the ER sorting signal to APP abolishes both AP40/42 and AP46 production whereas addition of the TGN sorting signal did not affect
Ap40/42 and AP46 generation ...... •...... •...... 13 4
Chapter 5 Fig.7. Co-localization of APPsw-ER with Grp78 and APPsw-TGN with y- adaptin ...... 13 5
lX LIST OF ABBREVATIONS
AD, Alzheimer disease AP, amyloid IJ-peptide ADAMlO, a disintegrin and metalloprotease-10 ADAMI 7, a disintegrin and metalloprotease-17 APP, P-amyloid precursor protein APPsw, Swedish mutant APP AICD, APP intracellular domain ApoE, apolipoprotein E BACE, P-site APP cleaving enzyme BF A, brefeldin A 31C, WPE-III-31C C99, P-secretase generated membrane bound APP fragment C83, a-secretase generated membrane bound APP fragment CTF, carboxy-terminal fragment CM, conditioned medium DAPT, N-[N-(3,5-difluorophenacetyl)-L-alanyl]-(S)-phenylglycinet-butyl ester D APM, N-[N-(3,5-difluorophenacetyl)-L-alanyl]-(S)-phenylglycine methyl ester FAD, familial Alzheimer's disease NCT, nicastrin Pen2, presenilin enhancer 2 PKC, protein kinase C PS, presenilin RNAi, RNA interference T ACE, tumor necrosis factora-converting enzyme TGN, trans-Golgi network
X CHAPTER! MOLECULAR BASIS OF ALZHEIMER'S DISEASE
In 1906, Alois Alzheimer, a German psychiatrist, reported at a meeting in Munich an unusual case of presenile dementia in which a 51-old woman, after suffering for several years from progressive memory impairment and cognitive disorder, eventually died.
Afterconducting an autopsy, Alzheimer found two distinct lesions in the patient's brain: senile plaques and neurofibrillary tangles, which distinguished the case from other diseases [1]. This is the first documented case of Alzheimer's disease (AD). Once recognized as a rare disorder in Alzheimer's time, with the increase of life expectancy,
AD has emerged as the most common form of dementia among the elderly. Over four million Americans and about 20 -30 million people worldwide are currently suffering from AD [2]. Unfortunately it will cause more serious health problems as the elderly population increases. Due to its distinct symptoms and growing number of patients, increased scientific efforts have been devoted to understanding the underlying mechanism of AD. Although it has been almost one hundred years since the initial report of AD, effective drugs to prevent this disease have not been found yet, and the disease remains incurable. However, much progress has been made toward the understanding of the clinical, pathological, genetic, cellular, and biochemical features of this most devastating neurodegenerative disease.
1 1.1 Some distinct features of AD
AD is clinically characterized by progressive memory loss, language deterioration, impaired judgment, and disorientation, and eventually results in global cognitive deficits.
The two distinct lesions first discovered by Alzheimer, intracellular neurofibrillary tangles (called NFT) and extracellular amyloid plaques (also known as senile plaques or neuritic plaques), have been regarded as two hallmarks of AD. In addition, microscope views also reveal a loss of neurons in certain regions of the brain such as hippocampus, a center for memory, and the cerebral cortex, where reasoning, memory, language and otherimportant cognitive actions take place.
Advances in biochemical pathology, immunocytochemistry and molecular genetics led to the identification of the main constituents of the two lesions. Neurofibrillary tangles are intraneuronal deposits containing paired helical filaments (PHFs) that are formed from the aggregation of hyperphosphorylated tau protein [3, 4]. The soluble form of tau containing 441 amino acids is one of the longest forms of unfolded native protein, lacking remarkable secondary structure[5]. The conformational change of tau from unfolded form into �-sheet structure leads to the highly ordered aggregates of PHF. In healthy neurons tau protein localized in axons of neurons binds to microtubules through its 3-4 repeated microtubule binding domains. The axonal microtubules functionas tracks for synaptic vesicles, helping transport neurotransmitters. However microtubules in neurons of AD patients collapse as a result of abnormal hyperphosphorylation of tau protein. Recent biochemical and animal model studies indicate that phosphorylation of
2 tau at its microtubule binding domain by microtubule affinity regulating kinase (MARK) under certain conditions dislodges tau from microtubules, thus leading to the further hyperphosphorylation of tau by GSK3 and Cdk5, which in turn triggers tau aggregation into filaments and tangles, leading to neuron dysfunction [6].
Amyloid plaques are generally found in large numbers in the limbic and association cortices [7]. The primarycomponents of amyloid plaques were isolated and sequenced by
Glenner and Wong in 1984 as small peptides with 39-43, predominately 40 and 42, amino acids[8, 9]. These small peptides designated as AP contain a high proportion of hydrophobic residues, thus adopting anti-P sheet conformation to form fibrils which deposits in plaques. AP42, a slightly longer, more hydrophobic form, is particularly prone to aggregation [10]. Amyloid plaques are surrounded by activated microglia, and reactive astrocytes. These fibril plaques are often accompanied by many "diffuse" (pre amy loid) plaques in the same brain regions which are thought to be precursors to amyloid plaques [2]. Interestingly diffuse plaques predominately contain AP42, indicating AP42 plays a pivotal role in the nucleation of plaques [ 11].
1.2 Hypotheses of the underlying mechanism of AD
Identification of major components in senile plaques and NFT provided clue for investigating the underlying mechanism of AD. In the search for the causes of AD, both tau tanglesand amyloid plaques have been suspected as the culprit of AD.
1. 2.1 Tau tangle theory. The tau tangle hypothesis emphasizes the role of tau tangle in the pathogenesis of AD. The finding that the number of NFTs correlates well with the severity of AD supports the tau tangle hypothesis. However, due to the later appearance
3 of NFTs in the brain of AD patient, and the evidence that transgenic mice over
expressing mutant tau did not trigger or enhance amyloid plaque formation, tau tangles
are now believed a downstream factor rather than the initial cause of AD.
1.2.2 Amyloid hypothesis. The amyloid hypothesis, also known as amyloid cascade hypothesis or AP hypothesis, was first proposed by D .J. Selkoe and J.A. Hardy more than a decade ago [12, 13] based on the following two lines of evidence: 1) the gene encoding p amyloid precursor protein (APP) is located in chromosome 21 [14-16]; 2)
Downs syndrome patients who have an extra copy of chromosome 21 invariably develop
the neuropathology of AD [17]. The amyloid hypothesis states that cerebral AP
accumulation is the primary influence in AD pathogenesis, and that the rest of the disease
process, including tau tangle formation, results from an imbalance between AP
production and AP clearance [18]. Although this hypothesis has not been universally
accepted, substantial evidence supports this theory. First, most of APP mutations in
familial early onset AD ( FAD) found so far are at or near the P-, y-, or a- cleavage sites,
and these mutations enhance either total AP production or the more toxic form of AP42
generation by favoring P- or y- secretase action [19-21]. Second, the cloning of the
presenilin (PS) proteins [22] and the discoveries that AD-causing mutations in PS 1 and
PS2 also enhance AP42 production reinforce the amyloid hypothesis [23]. Third, tau
mutations cause severe deposition of NFT and fatal neurodegeneration, but no amy loid
plaque formation[24] indicating that tau tangles can not trigger amyloid plaque formation.
Fourth, transgenic mice co-expressing mutant APP and mutant tau protein increase the
NFT formation, whereas the number of plaques remain unchanged compared to the
number of plaques in mutant APP transgenic mice, suggesting AP accumulation prior to
4 tau alteration in the pathogenesis of AD [25]. Fifth, injection of synthetic A�42 fibrils into the brain of tau P3 l OL mutation transgenic mice caused fivefold increases of NFT formation, providing directive evidence that A� fibrils can accelerate NFT formation in vivo [26]. Sixth, recent studies show that the genetic risk factor of late-onset AD, ApoE, may be involved in A� metabolism, because ApoE deficient mice expressing mutant
APP show significantly reduced cerebral A� deposition [27]. Finally, growing evidence supports the idea that genetic variability in A� clearance may contribute to the risk of late-onset AD [28]. Although the number of amyloid plaques does not correlate well with the severity of AD, furtherevidence indicates that memory andcognitive decline in mild AD patients correlate much better with cortical A� levels than plaque numbers, and correlate best with the soluble A� pool including soluble A� oligomers [29-31]. Recent evidence points to an important role of intraneuronal A� as a trigger of the pathological cascade of events leading to AD [32].
Taken together, biochemical, genetic and transgenic studies largely fit well with the amyloid hypothesis. The pathological cascade for the process of AD is likely to be: from amyloid accumulation and deposition to tau hyperphosphorylation and tangle formation and finallyto neuron death. Some therapeutic strategies based on the amy loid hypothesis aimed at reducing A� level such as using �- and y-secretase inhibitors are undergoing intensive investigations [33].
1.3 Genetic factorsin familial Alzheimer's disease
Epidemiology studies show there are two classes of AD, familial AD (FAD) and sporadic AD. It has been estimated that inherited formof AD accounts for 5-10% of total
5 AD cases [1]. However, FAD and sporadic AD are phenotypically highly similaror often indistinguishable, except for age of onset[l]. This finding strongly suggests that studies on FAD will provide important insights into the mechanism of sporadic AD. Currently
three genes, APP, presenilin 1 and presenilin 2, have been identified to cause autosomal
dominant early onset FAD (AD occurring before age 65), whereas e4 allele has been
found linked with late onset AD [34].
1.3.1 APP mutations. App gene mutations are estimated to account for up to 5% of
FAD [34]. The App gene, located on chromosome 21, encodes a protein called � Amy loid
precursor protein (APP), in which about a dozen missense mutations have been found to
be linked with early onset AD [16, 35]. Although these mutations have been found only
in about two dozen families worldwide, delineation of their genotype-to-phenotype
relationships have provided critical insights into the mechanism of AD. These mutations
are located either immediately before � cleavage site or near a-cleavage site, or shortly C
terminal to the y-secretase site (Fig 1. All tables and figures are located in Appendix),
especially at codon 71 7, where several different mutations have been identified. The two
mutations immediately before the �-cleavage site, also known as the Swedish mutation,
have been shown to favor �-secretase cleavage, thus increasing both A�40 and A�4 2
generation [20, 36], whereas mutations several amino acids C-terminal to the y-site have
been found to selectively elevate A�42 level, presumably by affecting y-secretase
cleavage [21, 35]. Over-expression of wild-type APP can also lead to early onset AD.
Due to an extra copy of chromosome 21 where the app gene resides, Down's syndrome
patients invariably develop clinical and pathological features of AD if they live over 30
years [37, 38]. They display diffuse plaques, which mainly contain A�2, in their teens
6 and 20s, and tangles a decade later. These findings indicate the critical role of A�, particularlyA�42, in thepathogenesis of AD [39].
1.3.2 Preseinlin mutations. Mutations in Presenilin 1 (PS1) gene (PSEN1) on chromosome 14q are the most causative factors in early onset FAD, which accounts for more than 60-70% of FAD cases. These mutations cause the most aggressive form of early onset FAD, commonly leading to onset of symptoms before the age of 50 and demise of the patient in his/her 60s. To date, more than one hundred mutations in PSl and nine mutations in presenilin 2, the homolog of PS1 which is located on chromosome lq31, have been identified [40]. PSENJ and PSEN2 encode 467 and 448 amino acid proteins, respectively, which are located primarily in the ER and Golgi apparatus. PSs comprise eight putative transmembrane (TM) domains. They undergo constitutive endoproteolysis to generate N-terminal and C-terminal fragments which bind together to form an active form [41]. Most mutations in PSs are missense mutations that are predominantly located in highly conserved transmembrane domains. Both transfected cells and transgenic mouse models have demonstrated that almost all of these mutations selectively elevate A�2 production [ 41-43], clearly indicating the essential role that presenilins play in y-secretase activity. Although PSs have been shown to function in protein trafficking and apoptosis, substantial evidence supports the idea that presenilins constitute the catalytic subunit of y-secretase (see y-secretase section). Thus mutations in
PSs directly affecty-secretase activity and specificity, generatingmore toxic forms of A�.
1.3.3 ApoE polymorphism. While mutations in APP and PSs lead to autosomal dominant inherited form of early onset AD, the t:4allele of Apolipoprotein E (ApoE) has
7 been found to be the major risk factor for late onset AD. Studies searchingfor proteins in human cerebrospinal fluid binding to immobilized A� led to the identification of ApoE, and the recognition that this gene is localized on chromosome 19q, a region which has
previously been found to link with some late onset AD cases [44]. Further genetic
analysis indicate that the frequency of e4 allele in AD patients is higher than that in the healthy population, and that while inheritance of one the e4 allele increase the likelihood
of developing AD by 2-5 fold, inheritance of two e4 alleles raise this risk by 4 to 10 fold
[45, 46]. In contrast to the elevated e4 frequencyin AD cases, the e2 allele frequency in
AD patients is low, suggesting ApoE2 may have some protective function against AD
[47].
Apolipoprotein E (ApoE) is a plasma glycoprotein containing 299 amino acids. Its N
terminal domain contains the receptor binding site, while the C-terminal domain has the
lipoprotein binding functions [ 48]. ApoE functions primarily in the mobilization and
redistribution of cholesterol during neuronal growth and after injury [ 49]. The
polymorphism of ApoE consists of single base changes at codons 112 and 158, which
encode Cysl 12 and Cysl58 in ApoE2, Cys112 and Asp158 in ApoE3, and Asp112 and
Asp158 in ApoE4 protein, respectively [50]. In contrast to the genetic variants in early
onset FAD, the presence of e4 is neither necessary nor sufficient to actually cause the
disease, suggesting that other genetic or environmental factors are required to cause AD.
The e4 allele has been shown to be strongly associated with increased neuritic plaques
and cerebral amyloid angiopathy in Alzheimer's disease [51]. Distinct binding properties
of different ApoE isoforms to A� and tau protein may provide a clue to explain the role
they play in AD pathogenesis. In vitro binding studies shows that ApoE binds to A� in an 8 isoform specific fashion, with ApoE4 binding more rapidly to A� compared to ApoE3
[52]. Specific binding of ApoE4 to A� may enhance A� aggregation or amyloid plaque formation [53]. Interestingly, ApoE3 and ApoE2 bind to tau protein more tightly than
ApoE4. This finding suggests that ApoE-tau complex may prevent tau from being phosphorylated, thus reducing tau-tangle formation[54].
1.3.4 Candidate susceptibility genes. Some other susceptibility genes have also been shown linkages with late onset AD, however these linkages have not been fully confirmed. a-2-Macroglobulin (a-2M) is a serum pan-protease inhibitor, also expressed in the brain, that has been implicated in AD on the basis of its ability to mediate the clearance and degradation of A� [55]. a-2M is also a component of senile plaques. Low density lipoprotein receptor-related protein (LRP) is the main receptor expressed in neurons. Genetic associations between two different LRP polymorphisms and AD have been observed [56]. Both the angiotensin converting enzyme (ACE) and insulin degrading enzyme (IDE) genes have been reported to be genetically linked with late onset AD, however these linkages have not been firmly confirmed [57, 58]. Recent results revealed their fu nctions in degrading A� (see section 1. 7).
1.4 Amyloid precursor protein processing and AP generation
1.4.1 APP. A� is derived from a large type I transmembrane protein, amyloid precursor protein (APP), by two proteases designated as �- and y - secretase. APP gene has been cloned and been localized in chromosome 21 [14-16]. It comprises three splicing isoforms:
APP 770, APP 751, and APP 695. They differin that both APP 770 and APP 751 contain
9 exon 7, which encodes a 56-amino acid motif homologous to a Kunitz-type senne protease inhibitor domain (KPI). Both APP770 and APP751 are widely expressed in noneuronal cells and also occurs in neurons, while neuronal cells express predominantly
APP695 [1]. All three APP isoforms have a signal peptide at the N-terminus, and a TM
domain near the C-terminus. APP undergoes extensive posttranslational modification
such as 0- and N- glycosylation, sulfation, and phosphorylation through the secretory
pathway. A� is generated by two sequential cleavages: �- site and 'Y- site cleavages. A
portion of APP is first cleaved at the �-site, releasing its large soluble ectodomain
fragment (�-APPs) into the lumen or extracellular space, resulting in a 99-residue C
terminal fragment (CTF99 or C99) which is still embedded in membrane. The resultant
C99 is furtherprocessed by 'Y-secretase at the "(-site, generating A�40 or A�42, and a C
terminal fragment(CTF'Y) [59]. Alternatively, APP is firstcleaved by a-secretase activity
near the middle of the A� region (between Lys16 and Leu17 of A� sequence), shedding
the long N-terminal ectodomain of APP (a-APPs), and yielding an 83-residue CTF83.
CTF83 is furtherprocessed by 'Y-secretase at the "(-site, generating P3 fragment (A� 17-
40/42) and CTF"f (60, 61]. Since P3 fragmentis not present in amyloid plaques, it seems
to be benign. Thus this process is termed as the nonamyloidogenic pathway, since it
precludes A� formation. Most APP molecules that undergo secretory processing are
cleaved by a-secretase rather than �-secretase, thus only a small amount of APP goes to
the amyloidogenic pathway [60, 61].
1.4.2 a-secretase: a-secretase is a putative protease that cleaves APP between Lys16 and
Leul 7 of A� sequence, generating a soluble APP exdodomain fragment (sAPPa) and
CTFa or CTF83. Although a-secretase has not been well defined, several candidate
10 proteins have been proposed. They are membrane bound ADAM (a disintegrin and metalloprotease) family proteases including ADAMl 7 (also called tumour necrosis factor a-(TNF-a-) converting enzyme or TACE )[62], ADAMl0 [63], and ADAM9 [64].
Currently it is unclear whether one or all of them constitute physiological a-secretsae.
Activation of PKC a-, P-, and e have been widely confirmed to increase a-secretase activity. However, deletion of the ADAMl 7 or the ADAMl0 gene has been shown to affecte only regulated form of a-secretase activity, suggesting that an unknown constitutive a-secretase exists [65]. Since the nonamyloidogenic pathway precludes AP formation, intensive effortshave been focused on modulating nonamyloidogenic pathway by increasing a-secretase activity[66]. However there are conflicting data regarding whether increased a-secretase activity by activation of some isoformsof PKC, such as a and e could decrease AP production [67-70].
1.4. 3 P-secretase. Among a-, P-, and y-secretase, P-secretase is the only secretase that has been well defined. P-secretase, also called P-site APP-cleavage enzyme (BACE), has been identified by several groups [71-73]. There are two homologues of P secretase,
BACEl and BACE2. The crystal structure of BACEl with its inhibitor has been determined at 1.9 angstrom resolution [74]. All these results firmly demonstrated that � secretase was a membrane bound novel aspartyl protease. BACEl mRNA is found in brain and other tissues, but it shows high activity only in brain [71]. Similarly to other proteases, BACEl is expressed as a pre-enzyme localized mainly in the· ER. The pre domain is cleaved by furin or other proprotein convertases [75] aftertrafficking through the Golgi. Over-expression BACEl in cultured cells increased P-cleavage activity [72], whereas a knockout of the BACE 1 gene in mice abolished A� generation,
11 demonstrating that BACEl is the maj or �-secretase in brain [76]. Purified BACEl exhibited high �-secretase activity, indicating that its activity does not require other cellular co-factors [73]. In addition to the major cleavage at the �-site, BACEl also cleaves at residues 10 and 11 of the A� sequence (referred to as W- site). Another transmembrane aspartyl protease, BACE2, which has 55% homology to BACEl, shows similar substrate specificity; however it is not highly expressed in the brain [73]. Thus
BACE2 is thought to play a less important role in A� generation. BACE 1 becomes a prime targetfor drug discovery in Alzheimer's disease, not only because it is the major � secretase in brain, but also because BACEl deficient mice were perfectly healthy, behaved normally, anddid not show any side-effects[7 7].
1.5 y- secretase complex
1.5.1 Presenilins. y-cleavage is the last step in A� generation, which determines the C terminal heterogeneity of A�. Since A�42 and A�43 are more hydrophobic, more prone to aggregation, and more toxic to neurons than A�40, the step which determines the ratio of A�42 to A�40 is crucial in the pathogenesis of AD. The other feature of y-secretase that made it more important is that y-cleavage occurs in the middle of a membrane, an extremely hydrophobic environment where normal enzymatic hydrolysis is impossible.
Thus elucidation of the mechanism of y-secretase cleavage is of great significance in AD prevention as well as in basic enzymology. Although we are still far from completely understanding it, much progress has been made toward elucidating the nature of y secretase.
Genetic studies have shown that mutations in PS 1 lead to the most aggressive form of
12 early onset FAD [78]. Mutations in presenilins account for most cases of FAD. More than one hundred mutations in PS 1 and eight mutations in presenilin2 (PS2) have been found to be associated with early onset FAD ( onset before age 65). Almost all of these mutations invariably increase the ratio of A�42 to A�40, demonstrating an essential role of presenilins in the y-cleavage [21].
The PSI gene was cloned a decade ago [22]. PSI is localized predominately in the ER and Golgi bodies. A portion of PS 1 has also been found in plasma membranes and endosomes. It has eight putative TM domains. PS 1 undergoes endoproteolytic processing within its largecytosolic Loop between TM6 and TM7, generating NTF and CTF, which remain bound together forming a stable, heterodimeric PS complex [79]. A major advance in PS 1 functional studies came from PS knockout mice. Primary neurons from
PS 1 deficient mice showed dramatic reduction in A� production when transfected with
APP [80]. PS 1 and PS2 double knockout cells neither altered APP maturation and distribution, nor affected a- and �-site cleavage processing, but completely abolished y secretase activity and A� production, leading to accumulation of C99 and C83, which are direct y-secretase substrates[81, 82]. These findings firmly demonstrate the essential role that presinilins play in y-secretase activity.
Another advance in understanding PS 1 function came from mutagenesis studies. PS 1 contains two conserved aspartate residues (Asp257 and Asp386), which are in TM6 and
TM7, respectively. Mutation of either of the two aspartate residues inhibited PS 1 endopreoteolytic processing, diminished A� production, and enhanced accumulation of
C99 and C83, a phenomenon similar to PS 1 knockout. Further studies show mutation of these two aspartate residues in a natural PS 1 splicing variant, PS 1�9 mutation that does
13 not undergo endoproteloysis but still functions, also blocked A� generation, confirming
that the blockage of y-secretase activity via mutation of the two conserved aspartate
residues is independent of PSI endoproteolysis [83]. These findings plus previous y
secretase inhibition results strongly suggest that y-secretase is an aspartyl protease, that
PS 1 is likely a catalytic subunit of y-secretase, and that the two conversed aspartate
residues likely form an active site of the enzyme. Further studies on PS2 also confirmthe
requirement of two conversed aspartateresidues [84-86].
More direct evidence that presenilines are catalytic components of y-secretase came from photoactive affinity labeling studies employing transition state analogue inhibitors to
target the active site of aspartyl protease. Photoactivable transition state analogues have
been found to covalently bind to both the N-terminal and C-terminal fragments of
presenilines, whereas none has been discovered bound to full-length presenilines which
are inactive forms [87, 88]. Interestingly, this inhibitor also has been shown to bind to the
PS 1 �E9 splicing mutant which is not endoproteolytically processed due to its lack of
endoproteolysis site, but is still an active formof presenilin.
Taken together, all these findings strongly support the idea that PS constitutes the
catalytic subunit of y-secretase. However, over-expression PSI did not enhance y
secretase activity, indicating that other cellular cofactors are required for y-secretase
activity. Indeed, y-secretase has been found in a high molecular weight complex (250kD
to 1000 kD). These findings spurred the searchfor other y-secretase cofactors.
1.5.2 Nicastrin. Using anti-PSI antibody to immunoprecipitate y-secretase complex,
nicastrin (N CT) was the first proteinto be purifiedand identifiedin the HMW y-secretase
complex. NCT is a highly glycosylated type I transmembrane protein with 709 amino
14 acids colocalized in the ER and Golgi apparatuses. Suppression of NCT expression with
RNAi in Caenorhabditis elegans resulted in a phenotype that is similar to those induced by the mutations of presenilin homologues in Caenorhabditis elegan. Co immunoprecipitation of NCT with full-length APP, C-99, and C-83 has been observed in several cell lines as well as in human brains. Mutations in a conserved hydrophilic domain ofNCT increased the ratio of A�42 to A�40, whereas deletion of this hydrophilic domain abolished the A� generation [89]. Genetic ablation of NCT in Caenorhabditis elegans and Drosophila resulted in a phenotype similar to the deletion of PS [90-92].
Fibroblast cells derived from NCT knockout mice completely abolished A� production, confirmingthe requirement ofNCT in the functionaly-secr etase complex [93].
1.5.3 Aphl and Pen-2. Genetic screening of genes in C. elegans that affect y-secretase activity by monitoring Notch signal pathway (another important pathway in which y secretase is involved) identified two novel proteins that are indispensable fory-s ecretase activity,Aph l [94] and Pen-2 [95]. Two homologue genes in human, Aphlla and Aphl b, have been identified, both of them encoding proteins that have seven putative transmembrane domains. Mutation of the conserved GxxxG motif in the fourth TM of
Aphl led to failure ofy-secretase assembly and loss of y-secretase activity, indicating its requirement for y-secretase complex activity [96]. Pen-2 (presenilin enhancer) has been identified as an enhancer of y-secretase activity in C elegans. It has been predicted as a transmembrane protein that spans the membrane for two times. Down regulation of this gene by RNAi technique reduced presenilin fragment accumulation and caused a loss of y-secretase activity.
Despite the fact that NCT, Aphl, and Pen-2 all affect y-secretase activity, individual
15 over-expression of each of them did not increase y-secretase activity, suggesting thatthey
work together to form a functional complex. Growing evidence shows that first Aph-1 and NCT form a subcomplex to bind to PS, then this subcomplex further binds to Pen-2
to initiate endoproteolysis of PS 1, thus conferring y-secretase activity [97]. Co
expression these fourprotein in yeast Saccharomyces cerevisiae, which lacks endogenous
y-secretase activity, successfully reconstituted y-secretase activity, further demonstrating
the necessity and sufficiency of these fourproteins in y-secretase complex [98].
In summary, substantial evidence demonstrates that presenilin, NCT, Aphl, and Pen-2
form a functional y-secretase complex, with presenilin as the catalytic subunit of the
multiprotein complex.
1.5.4 Other substrates of y -secretase. In addition to APP, y-secretase has also been
reported to cleave several other type I transmembrane proteins, including Notch[99],
Erb4[100], APLPl (APP Like protein-I), APLP2 (APP like protein2) [101], E-cadherin
[102], and CD44 [103] within their TM region (except for E-cadherin). Among these
substrates, Notch cleavages have been well studied. Notch is a type I transmembrane
receptor that is essential in determining cell fate during embryogenesis and in adults.
Upon ligand binding, Notch is first proteolytically processed by TACE [104] or ADAM
at the S2 site (similar to a-cleavage site in APP), which is 12 amino acid residues away
from the membrane[! 05], releasing its extracellular domain. The resultant membrane
bound fragment is cleaved by y-secretase activity at the S3 site (parallel to y-site in APP)
within its transmembraneregion [99], releasing Notch intracellular domain (NICD). Thus
the released NICD translocates into the nucleus, where it acts as a transcription
factor[l06, 107]. It has been found that PSl deficiency diminished NICD generation and
16 that y-secretase inhibitors also blocked S3 cleavage[99], suggesting that both Notch and
APP transmembrane cleavage depend on similar y-secretase activity[80, 99]. Indeed, recently a new cleavage site in the middle of the Notch TM domain has been foundand designated as the S4 site, which is similar to the y-site in APP[l 08].
1.5.5 &-cleavage: In addition to a-, �-, and y-site cleavages, a new secretase processing site, referred to as the e-site, has recently been identifiedby several labs at residue 49 of the A� sequence using a cell-free system [109-11 1]. A small portion of e-cleavage has also been shown at residue 52 of the A� sequence. e-sites are close to membrane boundary unlike the y-site that is at the middle of the membrane. This position is parallel to the S3 site in Notch. Another similarity between the S3 site and the e-site is that both cleavages have been shown to be PS dependency; and both can be blocked by y-secretase inhibitors. Similarly to NICD, the e-cleavage generated product, APP intracellular domain (ACID), has also been shown to translocate into nuclei to bind to Fe65 protein and Tiop60, acting as a transcription factor [1 12]. However, it is not clear what the relationship is between y-and e-cleavages, whether it consists of random chopping or sequential actions.
1.5.6 y-secretase Inhibitors. y-secretase inhibitors have served as useful molecular probes forstudying y-secretase. Screening protease inhibitors fory-se cretase activity has helped to identify y-secretase as a novel aspartyl protease. Since y-secretase plays a decisive role in A� generation, much effort has been focused on the search for ideal y secretase inhibitors. Based on APP sequence, some peptidomimetic inhibitors and their derivatives such as DAPT, DAPM,and compound E, have been designed and tested, and
17 exhibit strong inhibition. Since y-secretase is an aspartyl protease, aspartyl protease
transitionstate analogues, including L685458 and 31 C, which bind to the active site of y secretase, have been found. However, the progress of y-secretase inhibitor design has
been hampered by the observation that both S3 cleavage in Notch and e-cleavage in APP, which generate important transcription factors NICD and AICD, respectively, are also
inhibited by these y-secretase inhibitors. Some isocoumarin based inhibitors have been
reported to specifically inhibit A�42 generation without affecting NICD production [113].
However the subsequent work shows that the inhibition specificity seems controversial
[114]. Recently, some nonsteroidal anti-inflammatory drugs (NSAID) have been reported
to specifically inhibit A�42 production at high concentration [115, 116]. No doubt,
fu rther work needs to distinguish y-site cleavage from e-site and S3 site cleavages to get
ideal inhibitors which potently block A�40/42 production without affecting ACID ( e
cleavage) or NICD ( the S3 site in Notch) generation.
1.6 AP clearance
It has been widely accepted that the imbalance between A� generation and clearance is a
causative event in AD pathogenesis. Several peptidases have been reported to degrade A�
both in vitro and in vivo. A� clearance recently has emerged as the second potential
therapeutic approach to treat AD.
1.6.1 Neprilysin. Neprilysin is a zinc containing endopeptidase. It is an integral
membrane protein with a largeextracellular domain which comprises the catalytic site, a
TM domain which anchors the protein in the membrane, and a short intracellular domain.
Since its active site is outside of the membrane, neprilysin is the perfect enzyme for
18 digesting secreted peptides. Neprilysin primarily cleaves peptide bonds next to hydrophobic amino acids. It has been shown to degrade AP in vitro, mainly at Gly9 and
Tyrl O [117-119]. More direct evidence came from in vivo studies. After injection of radio-labeled AP peptides into rat brain, the labeled AP metabolized rapidly, especially
AP42. However, degradation of the injected AP was blocked by employing thiorphan or phosphoramidon, two neprilysin inhibitors. Interestingly, infusion of neprilysin specific inhibitor into the rat hippocampus led to the formation of "diffuse plaque". Moreover, deletion of neprilysin gene in mice resulted in elevated endogenous AP level, strongly suggesting that nerilysin plays a critical role in degrading AP42 and AP40 in the brain
[120]. Over-expression of this gene in cultured cell significantly decreased AP level [121].
Comparison of the neprilysin mRNA level and its protein level in brains from healthy individuals and AD patients shows inverse correlation with amyloid plaque number
[122], furtherconfirming the role of neprilysin in regulating AP level in humans.
1.6.2 Insulysin. Another enzyme, insulin degrading enzyme (IDE), was foundto regulate AP levels both in cultured cells [123, 124] and in vivo [125]. Insulin degrading enzyme
(IDE) is one of the major AP level regulators in neurons. IDE is a thiol zinc metalloendopeptidase located in the cytosol, endosomes, peroxisomes, and on the cell surface, degrading both insulin and Ap. IDE deficient mice showed increased cerebral cortex AP level. Elevated levels of the intracellular signaling domain of APP, which was recently found to be degraded by IDE in vitro, has also been found in IDE deficient cells
[125]. Moreover, genetic studies have found a linkage of the region of chromosomelOq, where IDE gene resides, with late-onset AD [57, 126]. However, later studies found another AP degrading enzyme, plasmin, which is also located in this region. It is hard to
19 attribute this linkage to IDE.
1. 6.3 Other enzymes. Angiotensin converting enzyme has also been found to degrade
A�. Interestingly, there is also a genetic linkage between the polymorphism· of this gene
andAD incidence[58, 127]. Other enzymes such as endothelin converting enzyme-1 [128]
and plasmin [129], have also been shown to have the capability of degrading A�.
1. 7 Conclusions and prospects
Alzheimer's disease is a devastating neurodegenerative disease, characterized by memory
loss and other cognitive dysfunctions. Among the two hallmarks, amyloid plaques and
their main component A� play a central role in the pathogenesis of AD. Three genes
linked with early onset AD all directly affect A�42/43 generation, whereas one gene
which is associated with late onset AD, ApoE, may be involved in A� clearance or
toxicity. �-cleavage enzymes have been cloned and well defined as membrane bound
aspartyl proteases. a-secretases have been suggested as metalloproteases and several
candidates have been proposed. The aspartyl protease nature and its four essential
components have been identified for y-secretase complex. Several genes involved in A�
degradation and clearance have also been identified. However, it is not clear what the
nature of y secretase is? What is the relationship between y-cleavage and £-cleavages?
Are the y- and the £-sites cleaved by the same secretase, or by similar but different
enzymes? This is an important issue for y-secretase inhibitor design aimed at reducing A�
without causing any side effects, because both AI CD and NI CD have important
physiological functions. Further work needs to address the following issues. How do
FAD mutations in APP and in PS affect £-cleavage? What is the role of intracellularA�?
How does A� affect tau tangle formation? How do the genetic mutations in the genes
20 involved in A� clearance affect A� accumulation? Addressing these issues will doubtlessly lead to new approaches to cure AD.
21 CHAPTER 2
IDENTIFICATION OF A NEW PRESENILIN-
DEPENDENT �-CLEAVAGE SITE WITHIN THE
TRANSMEMBRANE DOMAIN OF AMYLOID
PRECURSOR PROTEIN
This chapter is a slightly revised version of a paper by the same name published in the
Journal of Biological Chemistryin 2004 by Guojun Zhao, et al:
Zhao, G., et al., Identification of a new presenilin-dependent zeta-cleavage site within the transmembrane domain of amyloid precursor protein. J Biol Chem, 2004. 279(49): p. 50647-50. My use of "we" in this chapter refersto my co-authors and myself. I contributed most of the data and most of the writing.
Abstract y-secretase cleavage of �-amyloid precursor protein (APP) is crucial in the pathogenesis of Alzheimer's disease (AD), because it is the decisive step in the formation of the C terminus of �-amyloid protein (A�) and because the cleavage at A�42 is strongly linked to the disease. According to the amyloid hypothesis, the longer A� is more amyloidogenic and more pathogenic. Studies have also shown that the intracellular generation and accumulation of the long form of A� may account for the
22 neurodegenerative toxicity of A�. To better understand the molecular events involved in y-secretase cleavage of APP, in this study we report the identification of a new intracellular long A� species containing residues 1-46 (A�46), which led to the identificationof a novel l;-cleavage site between the known y- and s-cleavage sites within the transmembrane domain of APP. Our data clearly demonstrate that the new l;-cleavage is a presenilin-dependent event. It is also noted that the new l;-cleavage site at A�46 is the
APP717 mutation site. Furthermore, we show that the new l;-cleavage is inhibited by y secretase inhibitors known as transition state analogs but less affected by inhibitors known as non-transition state y-secretase inhibitors. Thus, the identification of A�46 establishes a system to determine the specificity or the preference of the known y secretase inhibitors, by examiningtheir effects on the formation or turnoverof A�46.
2.1 Introduction
In the brain of Alzheimer's disease (AD) patients, one of the characteristic pathological features is the deposition of amyloid protein in the parenchyma and in the walls of meningeal and cortical blood vessels. These amyloid deposits are principally composed of the 39-43 amino acid residue amyloid�-peptide (A�), which is derived from a large � amyloid precursor protein (APP). Now it is well established that, in the amyloidogenic pathway, APP is firstcleaved at the N-terminus of A� sequence by an enzyme named � secretase, to produce a soluble ectodomain, sAPP�, anda membraneanchored
C-terminal fragment, CTF�. CTF� is then subsequently cleaved within the transmembrane domain by another enzyme, referred to as "y-secretase" to produce the full-length A� and the intracellular domain (AICD) [1]). �-secretase has been identified
23 as a type I membrane aspartyl protease [72, 73]. The findings that knockout of presenilin
1 (PS1) and PS2 results in the abolishment of the y-secretase cleavage of APP and that two aspartate residues in two transmembrane domains of presenilin have been identified as critical for the y-secretase activity suggest that presenilin may be the y- secretase [80-
83]. Recently, several other molecules, namely, nicastrin, Aph-1, and Pen-2, have been identified as essential components of the "y-secretase" complex of which presenilin may function as the catalyticsubunit [130].
Most of the A� species contain 40 or 42 amino acids. Recently, sequence analysis revealed that the N-terminus of AICD starts at residue 50 of the A� sequence, which is 7-
9 amino acids away from the C-termini of A�40 and A�42. This led to the finding of the e-cleavage site between A�49 and A�50 [109-111, 131]. Now the cleavage at
A�40/42has been specifically referred to as y-cleavage site[11 O]. The identification of this e-cleavage site, which is equivalent to the S3 cleavage site in Notch 1, and the identification of the S4 site in Notch 1, which is equivalent to the y-cleavage site at A�40/42in APP, fu rther indicate that APP and Notch 1 undergo a similar intramembranous process, namely the y (S4) and e(S3) cleavages[108]. However, neither the intermediate A� peptide, which ends at the E cleavage site, nor the C-terminal fragment,which starts with an N-terminus generated by y-cleavage, has ever been detected. One possibility is that y- and e-cleavages occur simultaneously. The other possibility is that there may be additional cleavages(s) between y- and e-cleavages. Here we report that, in our effort to determine these possibilities, we successfully identified a novel A�-containing peptide in living cells. Using mass spectrometry, we determined this A� speciesas A�46 and this led to the identification of a
24 new cleavage site at AP46, which we designatedas ('.;-cleavage site following the tradition by which the APP processing sites have been named. It is notable that the newly identified('.;-cleavage site is the AD-linked APP717 mutation site.
2.2 Materials and Methods
2.2.1 Materials. y-secretase inhibitors, namely, y-secretase inhibitor IX (DAPT), y secretase inhibitor XVI y-secretase inhibitor XVIII (Compound E), y-secretase inhibitor
XIX (XIX), y-secretase inhibitor X (L-685,458) and y-secretase inhibitor XVII (WPE-III-
31C, 31C) were purchased from Calbiochem and dissolved in dimethy 1 sulfoxide
(DMSO). A�O was purchases from Biopeptide (California). AP42 and AP43 were purchased from American Peptide (California).
2.2.2 Cell lines. Human neuroblastoma Ml 7 cells, human glioma HS683 cells, and human embryonic kidney 293 cells were purchased fromAmerican TypeCell Collection.
All stably transfected N2a cell lines were established and maintained as described previously[132] . All cells were maintained in DMEM supplied with 10% FBS.
2.2.3 Immunoprecipitation and Western blotting. Eight hrs after treatment with inhibitors, cells were harvestedand lysed in Western blot lysis buffer (50 mM Tris-HCl, pH 6.8, 8 M urea, 5% P-mercaptoethanol, 2% SDS, and protease inhibitors). Secreted AP was immunoprecipitated from conditioned media using a AP-specific antibody 6E 10
(Senetek). Both cell lysates and the immunoprecipitates were analyzed by 10%
Bicine/urea SOS-PAGE or 10-18% regular SOS-PAGE. Aftertransferred to a polyvinylidene fluoride membrane (Millipore) and probed with specific antibodies, the immuno-reactivity bands were visualized using ECL-Plus (Amersham Biosciences).
25 2.2.4 Vector construction and transiently transfection. APPswN717F vector harboring both Swedish and Indiana mutations of APP were generated through site directed mutagenesis system (Stratagene), using Swedish mutant APP695 as a template.
Transiental transfection was performed uing Lipofectamine 2000 (Invitrogen) according to menufacture's instruction.
2.2.5 Mass spectrometric analyses. N2a cells cultured in the presence of DAPT were lysed with 1 % NP-40 in IP buffer (50 mM Tris/HCl, pH7.4, 150mM NaCl, 0.5% sodium deoxycholate, 5mM EDT A, protease inhibitor cocktail). After centrifugationat 20,000 xg at 4 °C for 15 min, the supernatant was diluted with equal amounts of IP buffer to bring down the concentration of NP-40 to 0.5%. The intracellular A� species were immunoprecipitated using 6E10. The immunoprecipitate was eluted frombeads with a buffer of 1 % TF A and 45% Acetonitrile and then subjected to Matrix-assisted laser desorption /ionization mass spectrometric (MALDI-MS) analysis performed on a
Voyager-DE STR mass spectrometer as described previously[l33].
2.3 Results
2.3.1 Identification of a new intracellular AP species. N2a cells stably expressing wild
type PSI (PSlwt) and the myc-tagged Swedish mutant APP (APPsw), which have been
used in previous studies [132, 134], were treated with or without y-secretase inhibitors:
DAPT, DAPM, compound E, and XIX. The cell lysates were analyzed by 10-18%
Tris/Glycine SOS-PAGE followed by Western blotting, using either 6El0, an
monoclonal antibody specific to residues 1-16 of human A�, or Cl 5, an antibody raised against the C-terminal 15 amino acids of APP. As shown in Fig. lA, in agreement with
26 previous reports, treatment with these inhibitors resulted in the accumulation of the
Cterminal fragments of APP, namely, CTF�, andCTFa, generated by �- and a-secretases, respectively, as detected by Cl5 (compare lanes 2, 3, 4, and 5 with lane1, top panel). The identities of these CTFs produced by this cell line have been described in a previous study[134]. Surprisingly, when the same samples were probed by 6El0, a band, which is tentatively labeled as A�46 (Fig 1 B, top panel lane 1) and has a migration rate much faster than that of CTF� but slightly slower than those of synthetic A�40, AP42, and A�43
(compare lane 1 with lanes 6 and 7, top panel Fig 1B), was detected in untreated cells when the blot was exposed to the film for prolonged time. More interestingly, this band was markedly increased in cells treated with y-secretase inhibitors ( compare lanes 2, 3, 4, and 5 with lane 1, top panel in FiglB). The factthat this band was not detected by C15, but detected by 6E10 suggests that it is more likely a novel intracellular A�-containing peptide with a higher molecular mass than AP43. The same samples were also analyzed by 10% Bicine/urea SDS-PAGE as described previously[135] . As shown in FigB, bottom panel, this new intracellular AP species (lanes 1 to 5) migrated much faster than
A�40/42and A�43 (lanes 6 and 7), indicating that this new A� speciesis more hydrophobic than AP43. Next, we examined the effects of these inhibitors on the formation of secreted AP.Secreted AP in conditioned media (CM) was immunoprecipitated with 6E10 and analyzed by Bicine/urea SDS-PAGE followed by
Western blotting using 6E10. As shown in the bottom panel of Fig. lA, secreted AP40 and A�42 were detected in CM of untreated cells (lane 1) but not detected in CM of cells treated with inhibitors(lanes 2 to 5), a result consistent with many previous studies.
To determine the physiological relevance of this new AP species, we examined its
27 presence in other types of cells. As shown in the top panel of Fig.2, in the presence of
3nM compound E, this new A� specieswas also detected in human embryonic kiyney
HEK293 cells( lane 2), human glioma HS683 cells (compare lane 4 with lane 3) and human neuroblastoma Ml 7 cell (compare lane 4 with lane 5) without overexpressing
APP. The new A� species migrated much faster than A�40/42(compare lanes 2, 4, 6 and
8 with lane 9) in a Bicine/urea SDS-PAGE. Since the blot was overexposed to visualize the new A� speciesproduced from endogenous APP, the densities of the bands corresponding to full-length APP (fA PP) and other non-specific or unknown bands were exaggerated. The same results were observed when DAPT, DAPM, and XIX were used
( data not shown). The detection of this new A� species in different types of cells indicates the physiological relevance of this peptide. Specifically, the detection of this new A� speciesin cells without overexpressing APP indicates that this new A� speciesis a normal metabolic product of APP.
2.3.2 Identification of a new �-cleavage site at AP46. The detection of the new intracellular long A�-containing peptide prompted us to determine whether it is the longsought intermediate A�-containing peptide, A�49, generated by e-cleavage. To determine the identity of these novel intermediate A�-containing peptides, mass spectrometric analysis was carried out. The new intracellular A� species was immunoprecipitated with 6El0 fromthe lysates of cells treated with DAPT, which causes increased accumulation of this peptide as observed in Fig 1. The immunoprecipitate complexes were analyzed by MALDI-MS. As shownin Fig. 3, a major spectralpeak with a mass of 4927 .52, which is in agreement with the expected mass of 4926 for A�46, was observed. Thus, the new A� specieswas identified as A�46. This result is in agreement
28 with the Western blot results that show that the new A� speciesis larger and more hydrophobic than A�43, suggesting that this new A� speciesis longer than A�43 at the hydrophobic C-terminus (Fig lB). Thus, we identified a new cleavage site at A�46 within the transmembrane domain of APP between the known y-cleavage site at A�40/42and the e-cleavage site at A�49. We named this new cleavage site as "�-cleavage" site, following the tradition by which other APP processing sites were named. It is noted that in addition to A�46, a weak spectral peak corresponding to A�43 (measured mass=4614.86; expected mass=4615) was detected in the immunoprecipitate.
2.3.3 �-cleavage is PSI-dependent. The novel finding that the new �-cleavage product A�46 was not inhibited, but rather increased by the known y-secretase inhibitors DAPT,
DAPM, compound E, and XIX prompted us to determine whether �-cleavage is dependent on presenilin. To this end, we examined the effect of the dominant negative aspartate mutant PS1 on the formation of A�46. In addition to the cells that stably express both APPsw and PS 1 wt used in the above experiments, another line of cells,
APPsw/PSl(D385A), which stably expresses both APPsw and the dominant negative mutant PSl(D385A) and has been used in a previous study [132], was employed. Cells were treatedwith or without 0.5 µM DAPT for 8 hours. The cell lysates and the secreted
A� immunoprecipitated from CM were analyzed by 10-18% SDS-P AGE followed by
Westernblotting using specific antibodies. As shown in the top panel of Fig.4 top panel, full-lengthPSl was detected in all cells transfectedwith APPsw and with either wild type
PSl or dominant negative PSl mutant. The PSl processing product, the N-terminal fragment of PS 1, was detected in cells expressing wild type PS 1 (lanes 1 and 2, top panel), but not in cells expressing the dominant negative PSl(D385A) mutant (lanes 3
29 and 4, top panel), which has been shown not to undergo endoproteolytic processing [83].
As shown in the middle panel of Fig.4, in the presence of DAPT, A�46 was detected in cells doubly transfected with APPsw and wild type PSI (lane 2) with concomitant decrease in secreted A� (bottom panel. Lane 2). Interestingly, A�46, as well as secreted
A�, was not detected in cells expressing dominant negative PS 1 mutant,regardless of the
presence or absence of DAPT (lanes 3 and 4), indicating that A�46 formation is PS 1-
dependent.
The PS dependence of A�46 formation was confirmed by using PS 1/2 double knockout
cells. Both PS1/PS2 wild-type and double knockout cells were transiently transfected
with either Lacz or APPsw695 and cultured in the presence or absence of compound E
for 12 h. As shown in Fig5A, A�46 was only observed in PS1/PS2 wild-type(lane8) but
not in double knockout cells(lane4) analyzed by both 10/18% SDS-PAGE(top panel) and
10% Bicine/urea SDS-PAGE(bottom panel), demonstrating the PS dependence of the
generation of A�46. As expected, secreted A�40/42were only detected from PS1/PS2
wild-type cells (Fig5B)
2.3.4 Transition state analogs inhibit �-cleavage. The data presented in Fig. 1 show that
the known y-secretase inhibitors DAPT, DAPM, compound E, and XIX, at the
concentrations used, completely inhibited the formation of secreted A�40 and A�42
produced by y-cleavage, but have less effect on the formation of A�46 produced by �
cleavage. This observation raises a possibility that these inhibitors may have differential
effects on y- and �-cleavages. In this regard, it was noted that the inhibitors tested,
namely DAPT, DAPM, compound E, and XIX are known as non-transition state
inhibitors. This prompted us to determine the effect of the inhibitors known as transition
30 state inhibitors on the new �-cleavage. We performed a series of dose-curve experiments to characterize these transition state and non-transition state y-secretase inhibitors. N2a cells expressing both APPsw and PS 1 wt were treated with various known y-secretase inhibitors at various concentrations. As shown in Fig.6, in cells treated with DAPT,
DAPM, compound E, and inhibitor XIX (lanes 10 to 17 and lanes 20 to 27), a dose dependent decrease in secreted A�40 and A�42 in CM (compare with samples from untreated control cells in lanes 9 and 19, lower panel) and, specifically at the lower range of concentrations of these inhibitors, a concomitant increase in CTF� and A�46 ( upper panel) were observed for all these inhibitors. Surprisingly, it was found that the newly identified A� speciesA�46 was not detected in the cells treated with transition state analogs, L-685,458 and 31 C (lanes 2 to 8, upper panel), though these inhibitors caused a marked and dose-dependent decrease in secreted A�40 and A�4 2 in CM ( compare lane 2 to 8 with lane 1, lower panel). This result reveals an important notion that s-cleavage, which is responsible for the formation of A�46, is specifically inhibited by inhibitors known as transition state analogs and is less affected by inhibitors knownas nontransition state inhibitors. All inhibitors examined at the ranges of concentrations used were found to have no effect on the cell growth and the expression and processing of PS 1 ( data not shown).
2.3.5 The Indiana mutation of APP enhances AP46 generation. Knowing that s cleavage site is the known V717 mutation site, we next examined the effect of Indiana mutation of APP on A�46 generation. As shown in fig 7 the level of A�46 generated from both APPswN717F is much higher(Fig 7 top panel lane4) than that of APPsw(Fig7 top panel lnae2), demonstrating that V717F enhances s-cleavage. Consistent with
31 previous reports, the ratio of A�42 to A�40 generated from APPswN717F (Fig7 lane3 bottompanel) is higher than that of APPsw (Fig7 lane 1 bottompanel).
2.4 Discussion
In this study we detected a new intracellular A�-containing peptide, A�46. The presence of A�46 in intact cells was confirmed in different types of cells. Specifically, the detection of A�46 in cells without over-expressing APP indicates that this new A� species is a normal metabolic product of APP. The finding ofthis new intracellular A�46 is also supported by recent observations that A� species ending at residue 46 can be detected in human tissue of subjects with and without AD[l36, 137]. It is noted that
A�46 exists at very low concentration in the absence of inhibitors and can only be detected by Western blotting after prolonged exposure of the blot, indicating its rapid turnoverin living cells.
The short life of this intermediate and its low concentration may account forthe difficulty in detecting it in living cells. In addition, the intracellular presence of A�46 may explain why it has never been detected previously in conditioned media. The identificationof this new A�46 species offers several novel insights into the mechanism of the normal and pathogenic intramembranous processing of APP and the formationof A�.
First, the identificationof A�46 reveals a novel cleavage site, [;-cleavagesite between the two known y- and a-cleavage sites. Our data also clearly reveal that this new [;-cleavage occurs as a presenilin-dependent event, indicating that it is a new intramembranous cleavage of APP by y-secretase or related protease activity. Moreover, the factsthat A�46 can be detected in the absence of inhibitors, and it is the predominant intracellular A�
32 speciesas determined by both Western blot and mass spectrometry analyses strongly indicate that the cleavage site at AP46 is another major cleavage site in APP, besides the known y-cleavage site at AP40/42and the e-cleavage site at AP49. It is also noted thatthe new �-cleavage site at AP46 is the site at which the AD-linked APP mutation, APP7I 7 mutation also known as London mutation, occurs[35, 138]. Thus, our finding reveals an interesting fact that the well characterized AD-linked APP7I 7 mutation actually occurs at a major APP processing site, the �-site near the C-terminus of Ap. Also the other well characterized AD-linked APP mutation, the Swedish mutation occurs at another major cleavage site, the P-site at the N-terminus of Ap.
Second, our dose curve experiments clearly demonstrate that, at the lower range of concentrations, DAPT, DAPM, compound E, and inhibitor XIX cause dose-dependent decrease in secreted AP40 and AP42 and concomitant increase in AP46, suggesting a possible precursor-product relationship between AP46 and AP40/42. This hypothesis is also supported by the fact that AP46 contains the y-cleavage site at AP40/42and that
AP46 is detectable in living cells in the absence of any inhibitors. However, the observation that �-cleavage can be differentially inhibited by y-secretase inhibitors known as transition state analogs, but less affected by the inhibitors known as non-transition state inhibitors also raises a possibility that �-cleavage may be catalyzed by a y-secretase like activity. And this y-secretase-like activity is also dependent on PS I or a PSI -like protein and can be specifically inhibited by transition state analogs, but it is different from the one that catalyzes the y-cleavage and can be inhibited by both transition state and non-transition state inhibitors. According to this model, the dose-dependent decrease in secreted AP40/42and the concomitant increase in AP46 caused by non-transition state
33 inhibitors at the lower range of concentrationsmay be related to the substrate availability, i.e. when the true y-secretase, which cleaves APP at A�40/42, is inhibited, it makes more
CTF� available for the putative y-secretase-like activity, which produces A�46. At higher ranges of concentrations, these inhibitors also caused dose-dependent increase of CTF�, suggesting an allosteric mechanism of the inhibitory effects of these inhibitors on the turnoverof A�46 and CTF�.
Third and more importantly, the identification of A�46 made it possible to determine the specificity of the known y-secretase inhibitors. To our knowledge, most of the known y secretase inhibitors have shown no difference in inhibiting the y- and E-cleavage as determined by their inhibitory effects on the generation of secreted A� and the APP intracellular domain(AICD), with the exception of isocoumarin "JLK", which remains controversial[l 13, 114]. However, our experiment examining the effects of the known inhibitors on the formation and turnover of the newly identified A�46 led to a novel finding that the newly identified s-cleavage is specifically inhibited by inhibitors known as transition state analogs and is less affected by inhibitors known as nontransition state inhibitors. According to the amyloid hypothesis [1 ], the longer and more hydrophobic and more amyloidogenic A� is more toxic and pathogenic. In this regard, it is notable that
DAPT, DAPM, compound E, and XIX, which were previously known to inhibit the formation of secreted A�, cause an intracellular accumulation of an even longer A� species, A�46 either produced by the same y-secretase, which also catalyzes the cleavage at A�40/42, or by a new y-secretase-like activity that specifically catalyzes the cleavage at A�46. This finding provides information important for the strategy of prevention and treatment of AD aimed at the design of y-secretase inhibitors. Therefore, the observation
34 that �-cleavage is differentially inhibited by so called transition state and is less affected by non-transition state inhibitors is significant because it establishes, for the first time, a system to determine the specificityor the preference of the known y- secretase inhibitors, by examining their effects on the formation or turnover of A�46. In general, the identification of the intracellular A�46 and the new �-cleavage site may provide a potential new therapeutic target forthe treatment and prevention of AD.
35 CHAPER 3
AMYLOID PRECURSOR PROTEIN IS PROCESSED
WITHIN ITS TRANSMEMBRANE DOMAIN BY
SEQUENTIAL E-, �-, AND y-CLEAV AGES
This chapter is about to be submitted by Guojun Zhao et al.
My use of "we" in this chapter refers to my co-authors and myself. I contributed most of the data and most of the writing.
Abstract
�-amyloid precursor protein (APP) apparently undergoes at least three major cleavages, y-, e-, and the newly identified s-cleavages, within its transmembrane domain. However, the roles of e- and s-cleavages in the formation of secreted A� and the relationship of these three cleavages, namely e-, s-, y-cleavages, remain elusive. We challenged these issues by attempting to determine the formation and turnover of the intermediates generated by these cleavages in the presence or absence of the known y-secretase inhibitors, using a differential inhibition strategy. Our data demonstrate that A�46 is the direct precursor of secreted A�. Furthermore, our data demonstrate that A�46 is associated with presenilin in intact cells. We also identified a long A� speciesthat is most likely the long-sought-after intermediate product, A�49, generated by e-cleavage and this
A�49 is further processed to generate A�46 and ultimately the secreted A�40/42. More interestingly, our data demonstrate that y-cleavage not only occurs last, but also depends
36 on �-cleavage and a-cleavage occurringprior to it. Thus, we conclude that the C-terminus of secreted A� is generated by a series of sequential cleavages, namely first a-cleavage followed by �- and y-cleavages.
3.1 Introduction
The mechanism of the formation of the �-amyloid protein (A�) is the central issue in
Alzheimer disease (AD) research, not only because A� is the major constituent of senile plaques, one of the neuropathological hallmarks in AD, but also because A� formation may be a causative event in the disease [1]. A� is proteolytically derived from a large single-transmembrane protein, the �-amyloid precursor protein (APP), as a result of the sequential cleavages by �- and y-secretases [1]. �-secretase has been identified as a type I membrane aspartyl protease [72, 73]. Though the exact nature of y-secretase is still a matter of debate, accumulating evidence supports the ideas that y-secretase is a multiple molecular complex composed of, at least, presenilins, nicastrin, Aph-1, and Pen-2 and that presenilin may functionas the catalytic subunit [130].
In understanding the mechanism by which the C-terminus of secreted A� is generated during the processing of APP, three major intramembranous cleavages have been established. The first one is the cleavage now specificallyreferred to as y-cleavage [11 O], which produces the C-termini of most of the secreted A� speciesthat end at amino acids
40 (A�40) or 42 (A�42) of the A� sequence. The second one is the a-cleavage occurring between A� residues 49 and 50, which produces the N-terminus of most of the APP intracellular domain (AICD) [109-111, 131]. The identification of this a-cleavage site raises a question as to whether this a-cleavage is obligatory for the generation of the C-
37 terminus of A�, and also raises a question as to the relationship between e- and y
cleavages, whether they are independent of each other or sequential. One of the obstacles
in addressing these questions is that neither the intermediate A� peptide, which ends at the e-cleavage site, nor the C-terminal fragment, which starts with an N-terminus
generated by y-cleavage, has ever been detected. In a very recent study, we reported the
identification of an intracellular long A� species, A�46. The identification of this long
A�46 led to the discovery of the third cleavage site, the s-cleavage site at A�46 between
the known y- andE-cl eavage sites [139] .However, the findingthat the known y-secretase
inhibitors, such as DAPT and compound E inhibit theformation of secreted A�40/42, and
on the other hand, cause the accumulation of A�46 raises a question as to whether
A�40/42and A�46 are produced by the same enzyme or produced by different enzymes.
Moreover, the roles of E- and l;-cleavages in the formation of secreted A� and the
relationship among these three cleavages, namely e-, l;-, and y-cleavages, also remain
elusive. To address these key issues, the objectives of this study were focused on: (a)
exploring the relationship between A�46 and A�40/42; (b) determining the possible
presence of A�49 generated by E-cleavage and its relationship with A�46 and A�40/42;
(c) examining the possible interaction of presenilin with A�46; (d) elucidating the
relationship of the three major cleavages, the E-, s-, and y-cleavages, within the
transmembranedomain of APP; and (e) establishing the roles of e- and l;-cleavages in the
formationof secreted A�40/42.
38 3.2 Materials and Methods
3.2.lMaterials. y-secretase inhibitors, DAPT, DAPM, compound E, L-685,458, and
WPE-III-31C (31C) were from Calbiochem and dissolved in dimethyl sulfoxide. A�40 and A�42 were purchased from American Peptide (California). A�46 and A�49 are customized peptides.
3.2.2 Cell lines and plasmids. N2a cells stably expressing wild type presenilin 1 (PS1 wt) alone or both PS1 wt and Swedish mutant APP (APPsw) were kindly provided by Drs.
Sangram S. Sisodia and Seong-Hun Kim (University of Chicago) and maintained as described previously [132]. The plasmid APPsw645, which expresses a C-terminal truncated APP ending at the e-cleavage site A�49, was constructed using the "Site
Directed mutaGenesis Kit" (Stratagene). APPsw [140], which was kindly provided by Dr.
Gopal Thinakaran (the University of Chicago, Chicago), was used as a template and a pair of oligonucleotides (E49: CGTCATCACCTTGTAGATGCTGAAGAAG; E49-r:
CTTCTTCAGCATCTACAAGGTGATGACG), which are complementary to each other and contain a stop codon at position 50 of the A� sequence, were used as primers.
3.2.3 Cell-free assay. In vitro turnover of A�46 by y-secretase activity was assayed in a cell-free assay system established previously [141], following the procedure described in ref. [109] with minor modifications. Briefly, N2a cells cultured in the absence of inhibitors were harvested in 9 volumes of homogenization buffer(10 mM MOPS, pH 7.0,
10 mM KCl) containing protease inhibitors (Complete, Roche) and homogenized by passing through a 20-gauge needle 30 times. After removal of nuclei and cell debris by centrifugation at 800 x g at 4 °C for 15 min, membranes were pelleted by centrifugation
39 at 20,000 x g at 4 °C for30 min. The membranes were washed once with homogenization
buffer and resuspended in assay buffer (150 mM sodium citrate pH 6.4, protease inhibitor
cocktail). Aliquots of equal amounts of membranes were then incubated at either 0 °C or
37 °C for the indicated time, in the presence or absence of y-secretase inhibitors. The
reaction was stopped by adding Laemmli SDS sample buffer containing 8 M urea. After
boiling for5 min, the samples were analyzedby Westernblotting.
3.2.4 Immunoprecipitation and Western blotting. Immunoprecipitation and Western
blotting were carried out as described previously[139] with the exception that some of
the immunoprecipitation was carried out in the presence of 500 nM DAPT as indicated.
Briefly, 24 hrs after splitting, cells were treated with inhibitors at various concentrations
or as a control, with vehicle only. Eight hrs after treatment, cells were harvested and lysed in Western blot lysis buffer (50 mM Tris-HCl, pH 6.8, 8 M urea, 5% P mercaptoethanol, 2% SDS, and protease inhibitors). Secreted AP was immunoprecipitated
from conditioned media using a monoclonal AP-specific antibody 6El0 (Senetek). To
immunoprecipitate the intracellular AP species, cells cultured in the presence of DAPT
were lysed with 1 % NP-40 in IP buffer (50 mM Tris/HCl, pH7.4, 150 mM NaCl, 0.5%
sodium deoxycholate, 5 mM EDTA, and protease inhibitor cocktail). Aftercentrifug ation
at 20,000 x g at 4 °C for 15 min, the supernatant was diluted with equal amounts of IP
buffer to lower the concentration of NP-40 to 0.5%. The intracellular AP species were
immunoprecipitated using 6E 10. Both cell lysates and the immunoprecipitates were
analyzed by 10% Bicine/urea SDS-PAGE or 10-18% regular SDS-PAGE and transferred
to a polyvinylidene fluoride membrane (Immobilon-P, Millipore). The membranes were
then probed with specific antibodies and the immunoreactivity bands were visualized 40 using ECL-Plus (AmershamBiosciences).
3.2.5 Fractionation and co-immunoprecipitation. In order to determine the formation of the possible complex of A�46 and presenilin, the following procedure, which was originally described in a previous study [142] and has been used to partially purify active y-secretase complex, was employed with slight modification. Briefly, N2a cells expressing APPsw695/PS1 wt cultured in the presence of 3 nM compound E (or 500 nM
L-685,458, Fig. SC) for10-12 hours were harvestedand homogenized in homogenization buffer A [20 mM HEPES, pH7.4, 50 mM KCl, 2 mM EGT A, 10% glycerol, protease inhibitor cocktail (Roche)] containing 10 nM compound E (or 2.5 µM L-685,458) by passing through a 20-gauge needle for 30 times. As illustrated in Fig. 4, the homogenized samples were subjected to centrifugation at 800 x g for 10 min to separate the unbroken cells and nuclei and the post nuclear supernatant (PNS). PNS was further centrifuged at
20,000 x g for 1 hour resulting in the supernatant (S 20k x g) and the pellet (P 20k x g).
The resultant pellet, which contains both A�46 and PSl (Fig. 5A), was solubilized in solubilization buffer B ( 50 mM PIPES, pH7 .0, 150 mM KCl, 5 mM MgC12, 5 mM CaCh, and protease inhibitor cocktail) [143] containing 1 % CHAPSO and 10 nM compound E
(or 2.5 µM L-685,458), for 1 hour at 4 °C, and then subjected to centrifugation again at
20,000 x g for 25 min to remove the insoluble materials. The supernatant was diluted with an equal volume of solubilization buffer B to adjust CHAPSO to a final concentration of 0.5%. After pre-clearing with Protein A sepharose beads for 3 hrs, the supernatant was incubated with anti-PSlN, a rabbit polyclonal antibodyraised against the
N-terminus of PS 1 [139] in the presence of compound E ( or L-685,458) with rotation at
4°C for 3-4 hours and then an appropriate amount of Protein A sepharose beads was 41 added and incubated overnight. As a control, an aliquot of the beads-pre-cleared supernatant was also incubated with pre-immune serum. After washing twice with solubilization buffer B containing 0.5% CHAPSO and y-secretase inhibitors and then twice with PBS, the immunocomplex was eluted with SDS-PAGE Sample loading buffer and separated by 10-18% SDS-PAGE forA�46 or 10% SDS-PAGE forPSl as described previously[139]. A�46 and CTF� were detected with 6E10. PSl was detected with anti
PSlL, which recognizes the C-terminus of PSl [144].
3.3 Results
3.3.1 L-685,458 inhibits the formation of Ap46. In our recent study, we have shown that treatment of cells with non-transition state y-secretase inhibitors, such as DAPT,
DAPM, and compound E caused an increase in the accumulation of intracellular A�46, indicating that these inhibitors have no or less effect on the newly identified l;-cleavage.
On the other hand, when the cells were cultured in the presence of transition state analogs, such as L-685,458 and 31 C, A�46 was not detectable, strongly suggesting that these inhibitors inhibit the l;-cleavage and block the formation of A�46 [139] However, it cannot be ruled out that these inhibitors may fail to block the turnover of A�46, resulting in A�46 vanishing. To address these issues, N2a cells expressing both PSl wt and APPsw were treated with DAPT, compound E, and L-685,458, either individually or in combination. Both cell lysates and the secreted A�40/42immunoprecipitated from conditioned medium (CM) were analyzed by 10% Bicine/urea-SDS-PAGE as described previously [135], followed by Western blotting using 6E10. Lane 15 is the mix of synthetic A�40 and A�42. As shown in Fig. 1, treatment with 0.5 µM DAPT (lane 2) or 5
42 nM compound E (lane 3) caused a marked decrease in secreted AP40/42(lower panel) and a concomitantintracellular accumulation of AP46 (upper panel) as has been shown in our recent study[139]. On the other hand, treatment with L-685,458, at a range of concentrations from 0.5 to 2.5 µM, completely abolished the formation of secreted
AP40/42in the conditioned medium (CM) (lanes 4-6, lower panel), whereas it did not cause the accumulation of intracellular AP46 (upper panel). To determine if the absence of AP46 was a result of the failureof L-685,458 to block the turnoverof AP46, cells were treated with L-685,458 plus 0.5 µM DAPT or plus 5 nM compound E, both of which have been shown to block the turnoverof AP46 [139], and see also lanes2 and 3, Fig. 1.
As shown in Fig. 1, the addition of DAPT (lanes 7 and 8) or compound E (lanes 9 and 10) did not lead to the accumulation of AP46 in the presence of L-685,458. This result clearly indicates that the absence of AP46 in cells treated with L-685,458 is due solely to its inhibition of the formation of AP46, rather than its failureto block the turnoverof Ap46.
3.3.2 L-685,458 does not prevent AP46 from turnover. Since the formation of AP46 was blocked in the first place, the experiments described in Fig. 1 did not address the question as to whether L-685,458 has the ability to block the turnover of AP46, in addition to inhibiting its formation. To address this issue, a system which contains preexisting AP46 is required. For this purpose, a cell-free system, which has been established and used in many previous studies to assay the in vitro y-secretase activity [109, 141], was employed. Cells were cultured in the absence of any inhibitors; the cell membranes were prepared, and the cell-free assay was performed in the presence or absence of inhibitors. As shown in Fig. 2, AP46 was detected in membranes prepared
43 from cells cultured in the absence of inhibitors followed by incubation at 0 °C for 1 hr
(lane 1 and lane 5, (lanes5-1 1 were second batch)). However, when the membranes were incubated at 3 7 °C for 1 h, A�46 became undetectable (lane 2 and lanes 6, 7), indicating the quick turnover of A�46 under this condition. Interestingly, when the membranes were incubated at 37 °C in the presence of either DAPT (lane 3 and 8) or compound E (lane 9), the amount of A�46 remained largely unchanged, indicating that DAPT and compound E inhibit the turnover ofA�46 in the cell-free system as they do in living cells. However, when the membranes were incubated at 37 °C in the presence of L-685,458, A�46 became undetectable (lanes 4 and lanes 10, 11). This result clearly indicates that, in contrast to DAPT and compound E, L-685,458 has no preventative effect on A�46 turnover. Taken together, the data presented in Fig. 1 and Fig. 2 show that L-685,458 specifically inhibits �-cleavage and blocks the formation of A�46, but has no effect on its turnover. A similar inhibition profile was also observed when 31 C was examined (data not shown).
3.3.3 AP46 is a direct precursor of AIJ40 and AIJ42. As reported in our recent study[139], at the low range of concentrations, DAPM, DAPT, and compound E cause a dose dependent decrease in secreted AP40/42 and a concomitant increase in AP46 (see also Fig. 1 ), suggesting a possible precursor-product relationship between AP46 and
AP40/42. This hypothesis is also supported by the factsthat AP46 contains the y-cleavage site at AP40/42and that A�46 is detectable in living cells in the absence of any inhibitors, suggesting that s-cleavage occurs prior to y-cleavage. However, it cannot be ruled out that these two cleavages may be unrelated events. The finding that the new s-cleavage
44 can be differentially inhibited by different inhibitors made it possible to determine the relationship between AJ346 and AJ340/42, and the role of the new �-cleavage in the formation of secreted A�40/42. As shown in Fig. 3, 24 hrs after splitting, cells were treated with either 100 nM DAPM (lanes 3 to 5 and 8 to 10; at this concentration DAPM completely blocks the formation of secreted A�40/42and causes marked accumulation of
AJ346, see Fig. 1) or with the vehicle dimethyl sulfoxide (Me2SO) (lanes1, 2, 6, and 7) for12 hrs. Then, L-685,458 (0.5 µM) was added to the cells in lanes 2 and 7. L-685,458 was also added to the cells in lanes 5 and 10, in addition to the existing DAPM, and continuously incubated for 40 min to completely stop the generation of new A�46 in these cells, because at this concentration, L-685,458 blocks the formation of A�46 (Fig.
1). Since L-685,458 has no effect on the turnover of A�46 (Fig. 2), this treatment also allows the complete turnover of the possibly existing A�46 in cells of lanes 2 and 7.
However, the existing A�46 in cells in lanes 5 and 10, accumulated during the last 12 hrs of culture, should remain unchanged because of the presence of DAPM, which prevents the turnover of A�46. As controls, cells in lanes 1 and 6 were untreated. All cells were then washed twice with fresh cold medium and cultured foran additional 1 hr (lanes 1 to
5) or 2 hr (lanes 6 to 10), either in the presence or in the absence of inhibitors as indicated.
Secreted A�40 and A�42 were immunoprecipitated using 6E10 fromCM of the last 1 hr or 2hr cultures. Both the cell lysates and the secreted A�40/42immunoprecipitated from
CM were analyzed by 10% Bicine/urea-SDS-PAGE followedby Western blotting using
6E10.
As shown in lanes 1 and 6 (lower panel, Fig. 3), secreted A�40/42was detected in cells cultured in the absence of inhibitors throughout the course of the experiment, and the 45 AP40/42is apparently produced in a time-dependent manner during the last 1 hr (lane 1) and 2 hrs (lane 6) of culture. As shown in lanes 2 and 7, CTFP was accumulated in a time-dependent manner, but neither secreted AP40/42nor intracellular Ap46 was detected in the cells treated with L-685,458 during the last two incubation periods (40 min and 1 hr or 2 hrs). As shown in lanes 3 and 8, no AP40/42was detected in the CM of cells treated with DAPM throughout the course of the experiment (lower panel). Instead, a concurrent accumulation of intracellular AP46 and CTFP was observed in these cells
(upper panel). In contrast, when DAPM was removed during the last 1hrand 2 hr incubation, the accumulated AP46 and CTFP were further cleaved to form secreted
AP40/42, in a time-dependent manner (compare lane 9 with lane 4 of both upper and lower panels). Interestingly, when DAPM was replaced with L685,458 during the last 1 hr and 2 hr incubation, the pre-accumulated AP46 decreased ( compare lane 10 with lane
5, upper panel), and secreted AP40/42concomitantly increased (compare lane 10 with lane 5, lower panel) in a time-dependent manner. It is notable that theaccumulated CTFP remained unchanged (compare lanes 10 with lane 5, upper panel). Thisresult indicates that L-685,458 has no inhibitory effect on the turnover of A(W6in living cells, which is in agreement with the result observed in cell-free systems (Fig. 2). The fact that the accumulated CTFP remains unchanged also indicates that the secreted AP40/42detected in these cells was solely produced from the pre-accumulated AP46. This conclusion is also supported by the observation that without pre-accumulating AP46, no secreted
AP40/42was detected in CM of cells cultured in the presence of L-685,458 during the last
1 hr and 2 hrs (lanes 2 and 7). Thus, these results clearly indicate that AP46 is the precursor of secreted AP40/42. The secreted AP40/42detected in lane 9 is the sum of the
46 AP40/42produced from both accumulated AP46 and CTFP, which was most likely converted to AP46 first,and then the resulting AP46 was further processed to AP40/42.
3.3.4 Ap46 is associated with PSl. Previous study has shown that as a substrate of y secretase, CTFP forms complex with PS 1, which is the putative catalytic subunit of the y secretase complex, at the sites of AP formation [145]. If AP46 is the precursor of
AP40/42, AP46, as an intermediate, may still be associated withy-secretase complex. To determine whether AP46 is still associated with y-secretase complex, a co immunoprecipitation experiment was performed as described in the Experimental
Procedures. As illustrated in Fig. 4, lysates of cells treated with compound E were first subjected to 800 x g centrifugationto remove the unbroken cells and nuclei (P 800 x g).
The resulting post nuclear supernatant (PNS) was subjected to further centrifugation at
20,000 x g resulting in the supernatant (S 20k x g), which contains the microsomal and cytosolic fractions, and the pellet (P 20k x g), which contains the crude plasma membrane and trans-Golgi network (TGN)-enriched fraction [146]. As shown in Fig. SA, AP46 was detected in the unbroken cell fraction of P(800 x g) (lane 1) and the plasma membrane and trans-Golgi network (TGN)-enriched fraction of P(20k x g) (lane 3), but not in the fraction of S(20k x g) (lane 2). Interestingly, as shown Fig. 5B, PS I was also detected in fractions of P(800 x g) (lane 1) and P(20k x g) (lane 3), but not in S(20k x g) (lane 2), indicating that AP46 co-fractionates with PSI into the 20,000 x g plasma membrane and
TON-enriched fraction. Therefore, as described in the experimental procedure, co immunoprecipitation was carried out after solublization of the P(20k x g) fraction using anti-PSlN, an antibody specific to the N-terminus of PSI [139]. As shown in Fig. SC,
47 A�46 was indeed co-immunoprecipitated with PSl from the crude TGN-enriched membranes prepared from cells treated with compound E (lane2), but not in cells treated with L-685,458 (lane 6), which inhibits the formation of A�46. In agreement with the previous study [145], CTF� was also co-immunoprecipitated with PSl (lanes 2 and 6).
The immunoprecipitates were also probed for the presence of PSl using anti-PSlL, an antibody specific to the C-terminus of PSl [144]. As shown in Fig. 5D, PSl is immunoprecipitated by anti-PSlN (lane 3) but not by preimmune serum (lane 2), indicating the specificityof the co-immunoprecipitation.
3.3.5 Detection of the possible Ap49. We also attempted to determine the possible presence of A�49 generated by e-cleavage. Lysates of cells cultured in the absence of inhibitor were subjected to immunoprecipitation followed by Western blotting using
6E 10. As shown in Fig. 6A, A�46 was immunoprecipitated from untreated cells (lane 7).
Interestingly, when the immunoprecipitation was carried out in the presence of DAPT, in addition to A�46, a band with a slower migration rate was detected (lane 8). Possibly due to the lower concentrationand the hydrophobicity, mass spectrometricanalysis of this A� specieswas unsuccessful. To estimate its molecular size, we synthesized two A� peptides,
A�46 and A�49. As it migrates at the same rate as that of the synthetic A�49 (lane 10), this new A� speciesis most likely the long-sought-after intermediate, A�49, generated by e-cleavage.
3.3. 7 AP49 is the precursor of AP46. To further determine the relationship between
A�49 and A�46, we created a construct, APPsw645, which expresses a C-terminal truncated APPsw ending at the e-cleavage site A�49. N2a cells, which stably express
48 wild type PS 1, were transiently transfected with APPsw645. As expected, AP49 was detected at a lower concentration (Fig. 6A, lane 4) in the absence of any inhibitors.
Interestingly, two bands were detected when the cells were cultured in the presence of compound E (Fig. 6A, lane 5). The lower band is apparently the AP46 (compare lane 5 with lanes 7, 8, and 9 of Fig. 6A). This result clearly indicates that AP49 is a precursor of
AP46. Interestingly, when these cells were treated with L-685,458, AP46 was not detected (lane 6), indicating that L-685,458 blocks the �-cleavage that produces AP46 fromAP49. In addition, the factthat the new band detected above the AP46 band in lane
8 migrates at a rate similar to that of recombinant AP49 (lanes 4 to 6) also provides further evidence supporting that this new AP speciesis the intermediate AP49 generated by e-cleavage at AP49. It should be noted that under normal conditions, AP49 can only be detected afterenrichment by immunoprecipitation carried out in the presence ofDAPT, indicating its rapid turnover. In the absence of inhibitors, the presence of AP49 in cells transiently transfected with APPsw645 may be due to the lower efficiency of (-secretase processing because of the lack of the C-terminal fragment of APP, which may be required to efficiently interact with y-secretase complex. As shown in lanes 1 to 3 of Fig.
6A, neither AP46 nor AP49 was detected in cells transfected with plasmid pcDNA3.1-
LacZ, which expresses Lacz, regardless of the presence or absence of inhibitors, indicating that the AP46 and AP49 detected in lanes 4 to 6 are not produced from the mouse endogenous APP, but are produced fromthe recombinant APPsw645.
We also endeavored to determine the presence of secreted AP40/42produced by
APPsw645-transfected cells. As shown in Fig. 6B, a band corresponding to AP40 was
49 detected in CM of the cells transiently transfected with APPsw645 in the absence of any inhibitors (lane 2), but not in the cells cultured either in the presence of compound E
(lane 3) or in the presence of L-685,458 (lane 4). This result is in agreement with the observation that these inhibitors cause intracellular accumulationof AP46 and AP49 (Fig.
6A, lanes 5 and 6). It was noted that the amount of secreted AP40/42produced by the transiently transfected cellsis very low. This is possibly due to the lower efficiencyof the processing of the C-terminal truncated APP expressed by APPsw645. Therefore, we attempted to establish stable cell lines that stably express this C-terminally truncated APP.
N2a cells, which stably express PSlwt, were transfected with APPsw645. Stable clones were selected by adding 0418 (for APPsw) andHygromycin (forpresenilin) to the media.
As shown in Fig. 6B, a relatively higher level of secreted AP40/42was detected in the
CM of cells stably transfected with APPsw645 (lanes 5 and 6, shown are two of the stable clones we selected). In support of our observation, the generation of secreted
AP40/42from the same C-terminal truncated APP mutant was also reported in a very recent study [14 7].
3.4 Discussion
By using the combination of L-685,458 and DAPT or compound E, we clearly demonstrated that the absence of AP46 in cells treated with L-685,458 is not due to its failureto block the turnoverof AP46, but, instead, is due exclusively to its inhibition of the formation of AP46. In addition, using a cell-freesystem, our data furtherdemonstrate that L-685,458 has no effect on the turnover ofAP46. A similar inhibition profile was also observed for 31C ( data not shown), indicating that these inhibitors, known as
50 transition state analogs, share the same inhibitory specificity, i. e. they specifically inhibit the formationof A�46 by ('.;-cleavage,but have no effecton the turnoverof A�46.
The observation that, under our current experimental condition, L-685,458 has no effect on the turnover of A�46 is important because this made it possible to determine the precursor and product relationship between A�46 and A�40/42 in intact cells. Using the differential inhibition approach, our data, in Fig. 3, clearly reveal the important finding that AP46 is the direct precursor of secreted AP40/42. This finding indicates that A�46 undergoes further y-cleavageto produce secreted A�40/42.
To further confirm the notion that A�46 is the direct precursor of secreted A�40/42, we performed the co-immunoprecipitation experiments and found that, as an intermediate product during the intramembranous processing by y-secretase, A�46 is indeed associated with PSl. In agreement with the previous study [145], CTF� was also co immunoprecipitated with PS1 (Fig. 5C, lanes 2 and 6). It was noted that in compound E treated cells, the amount of CTF� detected in the co-immunoprecipitate is less than that of A�46. It was also noted that in compound E-treated cells, a significant amount of
A�46 is co-immunoprecipitated with PSl, but A�49 was not detectable. These results may be explained by the possibility that once bound to presenilin, the initial substrate
CTF� is quickly cleaved at the E-cleavage site and the resulting A�49 is further and rapidly cleaved at the c'.;-cleavage site leading to the formation of A�46. The lower amount of the residue A�49 was not detectable under our current co-immunoprecipitation experimental condition. In the presence of compound E, the cleavage at y-cleavage site is inhibited and results in the accumulation of A�46. The accumulated A�46 occupies the
51 binding site of the y-secretase complex and prevents the further binding of CTF� to the y secretase complex, resulting in the accumulation of CTF� (Fig. 1, lanes 11 to 14).
However, most of the accumulated CTF� is not associated with PS1 andis detected in the fractioncontaining the microsomes and cytosol, which are freeof PS1 (Fig. 5A, lane 2).
Taken together, these results indicate that once bound to presenilin, the initial substrate,
CTF�, and the intermediate product, A�46, are closely associated withpresenilin until the release of thefinal product of secreted A�40/42. Thus, all the data presented support our hypothesis that secreted A�40/42is produced from APP by a series of sequential cleavages by y-secretase, i.e. e-cleavage at A�49, �-cleavage at A�46, and y-cleavage at
A�40/42.
The factthat L-685,458 and 31 C have no effect on the turnover of A�46 (Fig. 2) and the facts that these inhibitors block the formation of A�46 fromA�49 by s-cleavage (Fig. 6A) and that these inhibitors block the formation of AICD by E-cleavage [110, 111] indicate that transition state analogs L-685,458 and 31 C specifically inhibit E- and �-cleavages, but have no effect on y-cleavage. The finding that L-685,458 and 31 C do not directly inhibit y-cleavage strongly supports an important notion that these inhibitors inhibit the formationof secreted A�40/42in living cells by a mechanism other thandirect inhibition of y-cleavage, but indirectly, by inhibiting the formationof A�46. This idea is supported by the fact that A�46 is detectable in cells cultured in the absence of any inhibitor, indicating that �-cleavage must occur prior to y-cleavage, i. e. �-cleavage is upstream of y-cleavage. Therefore, the finding that inhibition of upstream e;-cleavage by L-685,458 and 31C completely prevents the downstream y-cleavage from taking place strongly
52 supports animpo rtant notion that, in living cells, y-cleavage not only occurs secondarily, but also is dependent on s-cleavage occurring first. In this regard, the fact that the putative AP49, which contains the s-cleavage site at AP46, is detectable in living cells in the absence of any inhibitors (lane 8, Fig. 6A) indicates that e-cleavage occurs prior to s cleavage, otherwisethe E-cleavage product AP49 should not have had a chance of being formed. Moreover, our data clearly demonstrated thatL-685, 458, which does not directly inhibit y-cleavage, also blocked the formation of secreted AP40/42from AP49 produced from the truncated APP mutant APP645 (Fig. 6B, lane 4), indicating that secreted
AP40/42 cannot be directly formed from AP49 without the formation of the intermediate
AP46 by s-cleavage. Thus, as illustrated in Fig. 7, our data strongly suggest that after P or a-cleavage of APP, the resulting CTFP and CTFa first undergo e-cleavage, followed by a sequential but rapid s-cle avage and then a y-cleavage, commencingat the site closest to the membrane boundary and proceeding towards the site within the middle of the transmembrane domain of APP. More importantly, y-cleavage not only occurs last, but also is dependent on E- and s-cleavages occurring prior to it. Accordingly, our data establish that the e-cleavage is not only an obligatory step, but also the initial step of the sequential s-, s-, and y-cleavages involved in the generation of the C-termini of secreted Ap.
The finding that E- and s-cleavages and y-cleavage can be differentially inhibited by transition state analogs and non-transition state inhibitors respectively, suggests several possibilities. First, these cleavages may be catalyzed by two enzymes and second, these cleavages may be catalyzed by one enzyme which has two inhibitor binding sites, one for
53 the transition state analogs, such as L-685,458, and the other for the non-transition state inhibitors, such as compound E, as suggested by a recent inhibitor binding kinetic study
[148]. The sequential relationship of these cleavages and specifically the finding that y cleavage is dependent on s- and �-cleavages occurring first strongly suggest that y-, c;-, and £-cleavages are catalyzed by a single enzyme. According to this model and the hypothesis described in ref. [148], the transition state analogs inhibit the initial cleavage, namely the £-cleavage, by directly binding to the catalytic site. As a result, the downstream c; - and y-cleavages are also prevented from taking place. On the other hand, the non-transition state inhibitors may bind to a remote site of the enzyme and induce conformational changes on the enzyme, resulting in the preferential inhibition of y cleavage with less effect on the E- and c;-cleavages by preventing the y-cleavage site from accessing the catalytic site of the enzyme. However, the one catalytic site model fails to account for the fact that AP46 is still processed to AP40/42by y-cleavage in the presence of the transition state analogs (Fig. 3), which is assumed to bind the catalytic site [148].
Therefore, the third possibility is likely that these sequential cleavages may be catalyzed by an enzyme which has two catalytic sites. Regardless of one or two catalytic sites, according to the one enzyme model, at high concentrations the non-transition state inhibitors, which preferentiallyinhibit y-cleavage, may also inhibit E- and c;-cleavages by an allosteric mechanism, i. e. these compounds may introduce conformational changes into the y-secretase complex, resulting in partial inhibition of s- and c;-cleavages. The one enzyme model is also supported by the fact that both groups of the inhibitors have been shown to bind to presenilins [88, 149, 150].
54 CHAPTER 4
CHARACTERIZATION OF y-SECRETASE
INHIBITORS: THEIR EFFECTS ON AfJ46
ACCUMULATION AND TURNOVER
This chapteris about to be submitted by Guojun Zhao et al.
My use of "we" in this chapter refersto my co-authors and myself. I contributed most of the data and most of the writing.
Abstract
Accumulation of AP, including intracellular AP, has been thought to play an early and crucial role in the pathogenesisof Alzheimer's disease. Inhibition of y-secretase is one of the promising therapeutic strategies to slow disease progression. Structurally diverse y secretase inhibitors have been demonstrated to abolish AP production both in cultured cells and in mouse models. However, most recently we demonstrated that application of some y-secretase inhibitors led to accumulation of intracellular AP46, which could be toxic to neurons. Here we have examined various y-secretase inhibitors regarding their effects on intracellular AP46 accumulation as well as AP46 turnover. We found that
NSAID y-secretase inhibitors ibuprofen and sulindac sulfide, and isocoumarin based inhibitor JLK6 did not lead to accumulation of intracellular AP46 or CTFP. On the contrary, high dose of NSAIDs blocked the endogenous level of AP46. Two new classes
55 of y secretase inhibitors, peptide aldehyde Z-Leu-Phe-CHO and fenchylamine sulfonamide based inhibitor (inhibitor VI) exhibited DAPT-like effect, leading to intracellular accumulation of both A�46 and CTF�, whereas peptide aldehyde 2-
Naphthoyl-VF-CHO displayed transition state analogue-like effects, accumulating CTF� withoutA�46. By examining the turnoverrate of pre-accumulated A�46 after removal of y-secretase inhibitor, we observed that benzodiazepine containing inhibitors displayed the strongest inhibitory stability. In the presence of L685,458, pre-accumulated A�46 was turned over into A�40 and A�42, while in the presence of 31 C, A�46 was processed into
A�43 and a longer A� species. Taken together, our data provide new insights into the understanding of various y-secretase inhibitors.
4.1 Introduction
Alzheimer's disease (AD) is a progressive neurodegenerative disease characterized pathologically by extracellular amyloid plaques, intracellular neurofibrillary tangles, and loss of neurons in the cerebral cortex[!]. According to the amyloid cascade hypothesis, the accumulation of �- amyloid protein (A�) plays an early and causal role in the pathogenesis of AD [12, 32, 138]. The primary components of amyloid plaques are heterogeneous A� species, predominantly A�40/42[8-1 O]. A�40 is the major species generated fromcells, whereas the slightly longer and less abundant formof A�42 is more hydrophobic, more prone to aggregation, and more pathogenic[lO]. A� peptides are generated from �-amyloid precursor protein (APP) by sequential actions of �- and y secretases. y-secretase has been shown to cleave APP in its transmembrane domain at two sites: the
56 y-site at A�40 or A�42 generating A�40/42, and the E-site at A�49 releasing APP intracellular domain (AICD) [109-111]. Most recently we identified another major y secretase cleavage site, the �-site, at A�46 in the transmembranedomain of APP[139]. In addition, y-secretase has also been identified to cleave several other type I transmembrane proteins such as Notch[99, 121], Erb4[100], CD44[103], E-cadherin[102],
APLPl and APLP2[101]. However it is currently unclear whether these cleavages are occured through the same or similar but slightly different mechanisms. This is an important issue forthe therapeuticapproaches to inhibit y-secretase activity.
Inhibition of A� generation by targeting either �- or y-secretases is one of the promising therapeutic strategies to prevent AD. However, it has been proven difficult to design small drug-like molecular inhibitors for �-secretase[l 51, 152]. Inhibition of y-secretase has also been hampered by the diverse substrates of y-secretase. Early y-secretase inhibitors were peptide aldehydes [153], or substrate based difluoroketones[154] and difluoro alcohols [155], which are generally weak inhibitors. Compound library screening and subsequent chemical optimization led to the developments of some potent semi-peptide inhibitors such as DAPT[156], compound E [150] and transition state analogues such as hydroxyethylene dipeptide isostere (L685,458) [157], and hydroxyethyl urea based 31C [149]. Although structurally diverse y-secretase inhibitors have been shown to effectively decrease A� level in cultured cells as well as in mouse models of AD[158, 159], most of these inhibitors also block other y-secretase mediated cleavages in addition to APP[l60]. Isocoumarin based inhibitor JLK6 has been reported to selectively reduce A� secretion without affecting other y-secretase mediated cleavages
[113]. However it failed to block solubilized y-secretase activity, indicating that it did
57 not act at the y-secretase level[114]. Recently, a subset of NSAIDs, such as ibuprofen, and sulindac sulfide, have been found to selectively lower AP42 level without affecting
AP40 and ACID generation [115]. However the mechanism by which NSAID drugs specifically inhibit AP42generation has not been well defined.
Most recently, we dissected APP transmembrane cleavages into three steps, and found that some y-secretase inhibitors at certain concentrations preferentially blocked y-site cleavage, causing intracellular accumulation of AP46[139]. Intracellular AP accumulation has long been speculated to play an early and important role in the pathogenesis of AD[161-16 3]. Growing evidence has demonstrated the cytotoxicity of intracellular AP [32, 164, 165]. It is of great importance to examine whether any other y secretase inhibitors could affect intracellular AP levels. Here we screened available y secretase inhibitors to assess their effects on AP46 accumulation as well as AP secretion.
4.2 Materials and Methods
4. 2.1 Materials. y-secretase inhibitors were all purchased from Calbiochem, and dissolved in dimethylsulfoxide (Me2SO). Monoclonal antibodies 22Cl 1 and 6El0, which recognize residues 66-81 of APP and residues 1-1 7 of the AP sequence, respectively, were from Boehringer Mannheim and Senetek, respectively. Poly clonal antibody C 15 raised against the C-terminal 15 residues of human APP has been described before [139].
HRP conjugated anti-rabbit and anti-mouse antibodies, Protein-A agarose beads, and
ECL-Plus westernblotting system were all obtained fromAmersham Biosciences.
4.2.2 Cell culture and inhibitor treatment. A mouse neuroblastoma cell line stably expressing Swedish mutant APP695 and human wild-type PS1 (N2aAPPsw695/PS 1 wt)
58 was established as described previously [132]. Cells were grown in Dulbecco's modified
Eagle's medium plus 10% fetal bovine serum for 24h. Then media were replaced with fresh media containing either Me2SO as a control or appropriate y-secretase inhibitors at the concentrations listed in Table 1. After incubation of cells with inhibitor for 10 h, conditioned media (CM) were collected and centrifuged at 3000x g for 10 min to remove cell debris. Cells were collected in Lysis buffer(50mM Tris-HCl, pH6.8, 2%SDS, 8M urea, 2mMEDTA, 2mM DTT, and protease inhibitor cocktail ) with a rubber policeman on ice as described previously[139]. After briefsonica tion, lysates were mixed with SDS
PAGE sample loading buffer and boiled for 5 min before loading. For the pre accumulated AP46 turnoverassay, N2aAPPsw/PS 1 wt cells were grown in the presence of
DAPM (Fig5) or other inhibitors as indicated (Fig4) for 12 h to accumulate AP46, then media were removed and cells were washed with cold media twice. Fresh media with or without inhibitor as indicated in each figure were added and the cells were further cultured for 2h before harvesting both the conditioned media and cells for analysis.
Appropriate y-secretase inhibitors were contained in wash media as indicated in figure legends.
4.2.3 Immunoprecipitation of AP from conditioned media. Immunoprecipitation of secreted AP from conditioned media was carried out as described previously[139] .
Briefly, centrifuged conditioned media were incubated with 6E 10 antibody for3 h at 4 °C followed by addition of protein-A conjugated agarose beads. After overnight incubation at 4 °C with shaking, the resulting immunocomplex-protein-A beads were collected by centrifugation at 5000x g and washed with 1 X PBS three times before adding SD S p AGE sample loading buffer and denaturing at 100°C for5min.
59 4. 2.4 Western blotting. Western blotting was performed as described previously [139].
For detection of soluble APP from conditioned media, samples were run on 8% SDS
P AGE Tris-glycine gels, transferred to polyvinylidene difluoride membranes, and probed with 22C 11 or 6E 10 antibodies. For analysis of intracellular A� speciesand APP fragments, samplesfrom cell lysates were separatedon a 10%-18% two-step Tris-glycine
SDS-PAGE system, transferredto polyvinylidene difluoridemembr anes, and probed with either 6E 10 or C 15 antibody. For detection of secreted A� species, immunoprecipitated samples were run on freshly made a 10% Bicine/Urea SDS-PAGE system, transferredto polyvinylidene difluoride membranes, and probed with 6E 10 antibody. All primary antibodies were incubated overnight and secondary antibodies incubated for 1 h. Bands were visualized withthe ECL-plus system (AmershamBio sciences).
4.3 Results
4. 3.1. Ibuprofen and Sulindac sulfide do not elevate intracellular AP46 level. The finding that some nontransition state analogues caused intracellular accumulation of
A�46 prompted us to examineother available y-secretase inhibitors. Recently, subsets of
NSAIDs, such as ibuprofen and sulindac sulfide, have been described to specifically block A�42 generation. To examine whether these NSAIDs affect intracellular A�46 level, we cultured N2a APPsw/PS 1 wt cells, which stably express Swedish mutant
APP695 plus human wild-type PS 1, in the presence of either ibuprofen, sulindac sulfide,
DAPT or compound E (as positive control) for 10 h. As expected, DAPT and compound
E treatment led to intracellular accumulation of A�46, as detected with 6E 10 from cell lysates( Fig lA, lanes 2-5). In contrast to DAPT or compound E treatment, we did not 60 observe an elevated level of A�46 in the lysates from either ibuprofen or sulindac sulfide treated cells (FiglA. top panel, Lanes6-8). On the contrary, 500µM ibuprofen and 60µM sulindac sulfide decreased endogenous level of A�46 (FiglA, bottom panel, longer exposure, compare lanes 7 and 8 with lane 1). Consistent with results from 6E 10 detection, no increased C-termini of APP were found fromibuprofen and sulindac sulfide treated cells when probed with Cl5 antibody (FiglB. compare Lanes6-8with lane 1).
These results indicate thatNSAIDs inhibit y-secretase through a mechanism distinct from that of DAPT. Indeed, NSAIDs have been shown to modulate rather than inhibit y secretase by switching the cutting site from A�42 to A�38[115], which may account for the lack of increase of intracellular A�46 level. Inhibition of endogenous level of A�46 by high concentration of ibuprofen and sulindac sulfide may be caused by nonspecific inhibition, since at this concentration A�40 level was also slightly reduced (Fig2C, comparelanes 18 and19 with lane10).
4.3.2. Peptide aldehydes and sulfonamide based inhibitors elevate intracellular AIJ46 level. To explore the effects of other y-secretase inhibitors on intracellular A�46 level, we cultured N2aAPPsw/PS1 wt cells in the presence of various y-secretase inhibitors and examined their inhibition profiles. In addition to DAPT, DAPM, and CPDE, which have previously been foundby us to increase A�46 level, we found two heavy bands in 1-(S) endo-N-(l ,3 ,3 trimethylbicyclo[2.2.l] hept-2-yl)-4-fluorophenyl sulfonamide
(inhibitorVI)[l 66] treated lysate (Fig2A, top panel lane6). The two bands are most likely
CTF� and A�46, respectively, as they ran at the same migration rates as those of DAPT treated samples in both 10-18% SDS-PAGE( Fig2A, top panel, lanes 6 and 7, for CTF� also see Fig2A, bottom panel, lanes 6 and 7) and 10% Bicine/Urea SDS-PAGE
61 system( data not shown), respectively. This finding indicates that in addition to peptidomimetic inhibitors such as DAPT, DAPM and Compound E, fenchylamine sulfonamidebased y-secretase inhibitor also causes increase of intracellular AP46 level. It was noted that although inhibitor VI decreased AP40 secretion, it did not alter secreted
AP42 level (Fig2C, compare lane 6 with lane 1 ). Since AP40 and 42 have been described as generated in different cellular compartments [167], variation in the ability of this inhibitor to reach different compartments may account forthe decreased inhibitory effect on AP42.
Application of peptide aldehyde inhibitor, N-benzyloxycarbonyl-Leu-phenylalaninal
(inhibitor V, or Z-Leu-Phe-CHO)[168], also led to accumulation of both AP46 (Fig2A, top panel lane5), and CTFP (Fig2A, bottom panel, lane5), though to a lesser extent when compared with DAPT treatment (Fig2A, both panels in lane5). Interestingly, although both AP46 and CTFP increased, treatment of inhibitor V at 25 µM did not block AP secretion, but rather dramatically elevated AP42 level and slightly increased AP40
( compare Lane 5 with Lane 1 in Fig2C), which is most likely due to the inhibition of non secretase-mediated degradation of CTFP[l53, 168]. In addition, it has been believed that not all CTFP undergoes secretase-mediated proteolysis, but rather is degraded[l 69].
We also found that treatment of naphthoyl-Val-Phe-CHO (Z-Val-Phe-CHO, or inhibitor
IV) caused dramatic increase of CTFP, but not AP46 (Fig2A, lane 4 in both panels), which is very similarto the L685, 458 and 31C induced effect(Fig2A, lanes 8 and 15, in both panels). Z-Val-Phe-CHO was originally identified as a potent cystein protease inhibitor[l 70]. It has also been shownto inhibit AP release in a dose dependent manner
[153]. Consistent with this previous finding, we observed a strong inhibition of AP
62 release (Fig2C, Compare lane 4 with lanel). Z-Val-Phe-CHO is not transition state analogue. If it inhibits y-, �-, and e-cleavages at equal potency, we should see either some longer AP species such as AP46 and AP49, or some CTFy and CTF�, in addition to CTFe.
Failure to detect other AP and C-terminal fragmentsin addition to CTFP suggests that Z
Val-Phe-CHO preferentially blocks the first cleavage among y-, �-, and e-sites, and cleavages of APP at its TM domain are sequential events.
Other peptide aldehyde inhibitors such as y40-I, y40-II, and XIV, did not exhibit significant accumulation of either AP46 or CTFP (Fig2A, lanes 2, 3, and 12). Inhibitors
XII, and XIII caused a slight accumulation of CTFP and even less amount of AP46
(Fig2A, lanesl0 and 11), presumably due to low dose of inhibitor. However, increasing concentration of these inhibitors resulted in a toxic effect as evidenced by cellular morphology change. Peptide aldehyde inhibitors have been reported to exhibit poor specificity[l71] , and this may explain why y40-I and XIV did not lead to accumulation of CTFs or Ap46 despite the reduced AP secretion(Fig2C lanes2 and 14).
For the isocoumarin derivative, although JLK6 moderately inhibited secretion of AP40 and 42(Fig2C, compare lanel1 with lanelO), no concomitant increase in the levels of
CTFs was found(Fig2A and B lane9). Further increasing the JLK6 concentration resulted in cytotoxicity in N2a cells. Previous observation suggested that JLK6 does not directly act on y-secretase, which may also account for the absence of elevated AP46 or
CTFP[114].
4.3.3 Inhibitor VI and 31C selectively abolish AP40 secretion. Having noticed that inhibitor V and VI selectively inhibit accumulation of AP40 over AP42, we adjusted the inhibitor concentrations to test whether other inhibitors could also display this effect. For
63 inhibitor V, no inhibitory effect was found at Sµm (Fig3A, lane4). Since we have observed enhanced secretion of A�42 at 25µm (Fig2C lane5), we increased the concentration to 75µm, however, this concentration led to some cytotoxic effects as noticed by cell morphology change. Interestingly, even at this high concentration, it still preferentially blocked A�40 over A�42 (Fig3B, lanes, longer exposure). Inhibitor VI preferentially blocked A�40 generation at 20µM (Fig3A, lane?), however, lower concentration did not elevate A�42 level (Fig3A, lane6). The transition state analogue
31 C specifically enhanced A�42 release at lower concentration (Fig3A, lane15), and had less effect on A�42 than on A�40 (Fig3A, lane15) at higher concentration. The other selected inhibitors, including another transition state analogue, L685,458, (Fig3A,lanesl 1 and 12) did not exhibit preferential inhibition with respect to A�40 and A�42.
Structurally diverse inhibitors including peptide aldehyde (inhibitorV), sulfonamind
(inhibitor VI), and hydroxyethyl urea (31 C) displayed preferential inhibition possibly via different mechanisms such as inhibitor accessibility, blocking nonspecific degradation, andallosteric effect.
4.3.4. Benzodiazepine containing inhibitors form the most stable complex with y secretase. Several inhibitors have been demonstrated to bind to y-secretase in a noncompetitive manner with substrate[148]. To assess the binding stability of inhibitor-y secretase complex, we firstcultured cells in the presence of y-secretase inhibitors for12 h, then removed inhibitors by washing with cold fresh medium twice, followed by adding fresh media without any inhibitor (for each corresponding control, fresh media with corresponding inhibitor), and incubated the cells for 2h. We found that removal of inhibitors IV, VI, XVI, and 31 C (XVII) led to elevated level of A� ( Fig4, top panel,
64 lanes 3,7, 14,16, respectively), with concomitant disappearance of pre-accumulated CTF� and A�46( Fig4,bottom panel, compare lanes3,7,14, and 16 with lanes 2,6, 13, and 15, respectively), suggesting that the binding of these inhibitor with y-secretase is unstable.
In contrast to the above results, pre-accumulated A�46 in lysates from inhibitor XVIII
( compound E) and XIX treated cells remained largely unchanged(Fig4, lower panel, comparelanes 18 and 20 to lanes 1 7 and 19), which suggests stable binding between these two inhibitors and y-secretase. Both inhibitor compound E and inhibitor XIX are benzodiazepine based drugs, and their structure may contribute to the strong binding.
Indeed these two benzodiazepinederivatives exhibited the lowest IC50 for A� generation.
4.3.5. L685, 458 and 31C differentially affect y-secretase specificity. Both L685,458
(inhibitorX) and 31C(inhibitor XVII) are aspartyl protease transition state analogues.
They both diminished A� secretion and concomitantly led to accumulation of CTF�.
Since L685,458 has been shownto display no inhibitory effect on A�46 turnover ( chapter
3 ), we attempted to examine whether and how 31 C affects pre-accumulated A�46 turnover. To accumulate A�46, cells were cultured in the presence of DAPM for 12h, then the media were removed and cells were washed with cold fresh medium twice. To block full length APP and CTF� processing into A�40/42, we employed L685,458, and
31 C, which both strongly inhibit A� generation from CTF�. Appropriate controls such as without pre-accumulation of A�46 and continuous incubation with DAPM and L685,458 were also included. Thirty-five min pretreatment forL685 ,458 and 31C was performed before washing step to allow these inhibitors to enter cells. Appropriate 'Y secretase inhibitors as listed in the figurewere included in washing media. After wash, fresh media with or without L685, 458 or 31 C were added followed by incubation for 2h before
65 harvesting conditioned media and cells. As shown in Fig 5 lanes 2, 3 and 6 to 9, both
L685,458 and 31C blocked CTF� turnover, but failed to inhibit A�46 turnover (Fig5 top panel, compare lanes 6 , 8 and 9 with lane4). Concomitantly, A�40 and A�42 bands were observed in samples from DAPM pre-treated cells subsequently cultured in the presence of either L685,458 alone or L685,458 plus Z-VAD (Fig5, lower panel, lanes 6 and 9), but not in the samples from L685,458 treated cells without DAPM pretreatment(Fig5, bottom panel,lane2), indicating that L685,458 did not affect A�46 turnoverintoA�40and A�42 under the test conditions. Adding Z-V AD to the medium did not block A�6 turnover ( compare lane 9 to lane 6, in both panel), which excludes the possibility that A�46 was turned into Abea40/42 via caspases activity. Intriguingly, two
A� bands with faster mobility than those of A�40/42were found in the sample from
DAPM pre-treated and subsequently 31 C treated cells(Fig5, bottom panel, lane8) but not from cells treated with 31C without DAPM pre-treatment (Fig5,lower panel, lane3), indicating that the two A� species were derived from A�46 in the presence of 31 C. Due to the lower amount, the identities of these two A� species have not been determined by mass spectrometry analysis. However, these two bands ran faster than A�40 and 42 in this Bicine/Urea SDS-PAGE system, suggesting that they contain more hydrophobic residues. As both 31C and DAPM do not alter �- or a-secretases specificity, these bands must start from first residue of A�. Thus, the major band is likely A�43 (Fig5, lower panel, lane8, upper band) as A�43 runs slightly faster than A�42 in this separation system [139]. The lower band in Fig 5 lane 8 is between A�44 and A�46. Similar results were obtained when using inhibitor VI to pre-accumulate A�6, whereas co-incubation of
31C either with DAPM or with inhibitor VI abolished any A� secretion ( data not shown).
66 The reason why AP46 turnover into AP40and 42 was seen in the presence of L685,458, whereas turnover into AP43 and 44/45/46 in the presence of 31 C was seen is not clear at present. Inhibitor structure differencecould account for the discrepancy in AP46 turnover.
Binding of 31 C to y-secretase may induce conformational change which alters y-secretase specificity, inhibiting y-secretase cleavage at 40 and 42 sites but favoring cleavages at 43 and 44/45 sites, whereas the conformation change introduced by binding of L685,458 differs from that induced by 31 C, as the structures of 31 C and L685,458 are different.
Indeed 31 C exhibited a less inhibitory effect on AP42 than on AP40 (Fig3A, lanes 15 and
16), while L685,458 displayed an equal effect on both AP40 and AP42 (Fig3A, lanes 11 and 12) secretion.
4.4. Discussion
Accumulation of AP is thought to be the central event in the pathogenesis of AD.
Targeting y-secretase is one of the potential therapeutic strategies in seeking to slow disease progression. Structurally diverse y-secretase inhibitors have been shown to diminish AP generation both in cells and in mouse models. Most recently, we have demonstrated that treatment of cells with some semipeptide inhibitors such as DAPT,
DAPM, or compound E elevated intracellular AP46 levels, whereas application of some transition state analogues including L685,458 and 31 C led to accumulation of CTF� without AP46. Growing evidence has demonstrated that accumulation of intracellular
A�42is toxic to neurons [164, 165]. It is conceivable that the longer and more hydrophobic AP46 could be more toxic to neurons than AP42. Thus it is necessary to re evaluate available y-secretase inhibitors with respect to their effects on intracellular AP speciesand CTFs, which is not only important for secretase inhibitor design, but also 67 facilitates understanding the underlying mechanism of y-secretase mediated cleavages.
For the present study we screened available y-secretase inhibitors to examinewhether and how these compounds could affect intracellularA�46 formation as well as A�46 turnover.
Here we foundthat two NSAIDs (ibuprofen, sulindac acid), one isocoumarin derivative
(JLK6), and some peptide aldehydes (y40-I, y40-II, XIV) did not increase accumulation of A�46 or CTF�. On the contrary, high concentration of ibuprofen and sulindac sulfide abolished endogenous level of A�46. Both ibuprofen and sulindac sulfide have been demonstrated to specifically block A�42 release without affecting A�40 and e-cleavage.
It has been suggested that these NSAIDs modulate y-secretase specificity, switching the cleavage site from A�42 to A�38, but largely do not affect A�40 cleavage [1 15].
Consistent with these previous observations, our result shows that these two NSAIDs exhibited distinct effects from those of peptidomimetic inhibitors and transition state analogues with respect to A�46 and CTF� accumulation. As intracellular A� species has been believed to be toxic to cells, reduction of A�42 secretion by some NSAIDs without increased intracellular A� species is of great significance for their potential therapeutic applications. JLK6 has been described selectively inhibiting A� release without affecting e-cleavage[l 13], however the mechanism remains unclear. Subsequent study has shown that JLK6 does not directly act at the y-secretase level[114], which could account forthe failure to detect elevated level of A�46 or CTF�. The reason why some peptide aldehyde inhibitors did not cause dramatic accumulation of A�46 and CTF� is most likely that they are weak inhibitors with low specificity. Due to nonspecific inhibition, increasing concentration leads to cytotoxic effects, resulting in low A� secretion. Thus reduction in
A� level does not cause significant accumulation of A�46 or CTF�.
68 Our previous studies have demonstrated that peptidomimetic and benzodiazepine containing inhibitors cause intracellular accumulation of both AP46 and CTFP. Here we report that two new classes of inhibitor, a fenchylamine sulfonamidecontaining inhibitor
(inhibitor VI), and a peptide aldehyde inhibitor (inhibitor V). Both exhibited effects similar to that of DAPT and compound E, although the aldehyde inhibitor displayed less potency. The observation that structurally diverse inhibitors display similar effects with respect to accumulation of intracellular AP46 strongly supports our previous finding that
�-cleavage is a physiological event closely related to y-site cleavage. Intriguingly, another peptide aldehyde, inhibitor IV (2-Naphthoyl-VF-CHO), mainly caused accumulation of
CTFP without accumulation of AP46, a result very similar to that produced by transition state analogues. If inhibitor IV blocks y- , �- and e-sites cleavages with equal potency, we should see some intracellular AP speciessuch as AP46 and AP49, at least at low concentration. The failureto see AP46 and AP49 suggests that APP C-terminal cleavages are not random, but rather sequential actions. Inhibitor IV most likely blocks the initial cutting site, thus leading to accumulation of CTFP alone.
Among the tested inhibitors, inhibitor V, VI, and 31C showed less potency for AP42 than
AP40. In particularly, inhibitor V and 31 C specifically elevated AP42 level at certain concentrations. Several factors could account for these results. First, binding of inhibitor to y-secretase induces conformational change, which may favor AP42 cleavage. Second, inhibitor accessibility may confine inhibitory effects to certain subcellular compartments where AP40 is generated, as AP40 and AP42 have been suggested to be generated in different compartments [ 167] . Third, moderate inhibition of AP40 cleavage at low inhibitor concentration leaves more substrates for AP42 cleavage, resulting in more AP42.
69 Finally, blocking nonspecific degradation may also contribute to the increase of A�42, since not all CTF� undergoes secretase mediated proteolysis[169]. It is conceivable that thesefactors may coexist in some cases.
Several inhibitors have been demonstrated to bind to y-secretase in a non-competitive manner [148]. In the present study, we analyzed inhibitor-y-secretase complex stability after removal of inhibitor from medium by evaluating the turnover of pre-accumulated
A�46 and CTF�. We found benzodiazepine-containing inhibitors displayed the strongest inhibitory stability. Inhibitor-y-secretase complex stability should depend on inhibitor concentration, binding affinity, and binding manner. Since the concentration of benzodiazepine containing inhibitors is the lowest among tested inhibitors, their strong binding comes fromeither high affinity or stable chemical bonds between inhibitor and y secretase.
Although both L685,458 and 31C are transition state analogues, which are believed to bind to the active site of y-secretase leading to accumulation of CTF�, differences between L685,458 and 31C have been noted in this study. First, L685,458 equally blocked both A�40 and A�42 generation while 31C displayed less potency for A�42 than forA�40. Second, A�46 was converted into A�40 andA�42 in the presence of L685,458, whereas it was processed into A�43 anda longer A� species in the presence of 31C. The reason for these discrepancies is not clear at present. One possible explanation is that binding of hydroxyethylene dipeptide isostere L685,458 to the active site of y-secretase may do not affect y-secretase specificity, whereas binding of (hydroxyethyl) urea based
31C to the active site of y-secretase most likely introduces a conformational change, which alters y-secretase specificity.
70 Taken together, our data provide new insights into the understating of various secretase inhibitors. Previous y-secretase screenings were focused on theireff ects on A� secretion as well as on e- and S3-site cleavages in the Notch signal pathway. As most tested y secretase inhibitors lead to either accumulation of CTF�, or CTF� plus A�46, which is believed to be toxic, further work on y-secretase inhibitors should also consider intracellular A� species.
71 CHAPTER S
Ap46 IS GENERATED IN THE GOLGI/TRANS
GOLGI NETWORK IN NEURO 2A CELLS
This chapter is about to be submitted by Ouojun Zhao et al.
My use of "we" in this chapter refers to my co-authors andmyself . I contributed most of the data andmost of the writing.
Abstract
Subcellular sites for generation of A� have been the subject of intensive research.
DifferentA� species are know to be generated in distinct subcellular compartments. We previously identified a novel cleavage, the �-site, in the transmembrane region of � amyloid precursor protein. Here we report that A�46 is generated in the Oolgi/trans
Oolgi-network (TON) compartments in neruo 2a cells. Using brief centrifugation, we found that A�46 was present in 20,000 x g membrane fraction but not in 20,000 x g supernatant. Cell surface trypsinization demonstrates that the majority of AP46 is resistant to trypsinization. Treatment with brefeldin A (BF A) completely abolished AP46 generation, whereas monensin only exhibited inhibitory effect on AP46 generation at high concentrations, suggesting that A�46 is generated in the Oolgi andTON. In addition, neither 25mM NH4Cl nor 1 OµM lactacystin affected A�46 level, indicating that A�46 is less likely generated in endosomes/lysosomes, and preteasomes. Furthermore, addition of an ER sorting signal to the C-terminus of APP diminished A�46 and A�40/42generation, whereas retention of APP in the TON with a TON sorting motif did not after A�46 or
72 AP40/42level, confirming that AP46 is generated in the Golgi/TGN but not in the ER.
Interestingly, we found that AP46 and AP40/ 4 2were differently affected by monensin,
NH4Cl, lactacystin, and tunicamycin. While these drugs displayed no or less inhibitory effects on AP46 generation, they significantly blocked AP40/42generation, suggesting that AP46 is likely generated at the earlier stage in the secretory pathway than AP40/42.
5.1 Introduction
Alzheimer's disease (AD) is a devastating neurodegenerative disease characterized by extracellular deposition of amyloid plaques and intracellular formation of neurofibrillary tangles. Overwhelming evidence supports the hypothesis that accumulation of AP, the predominant component of amyloid plaques, plays a crucial role in the pathogenesis of
AD[l, 18]. AP is derived from itsprecursor protein, a type I transmembrane(TM) protein called P-amyloid precursor protein (APP) through sequential cleavages by P- and y secretases. P-secretase librates the large N-terminal ectodomain of APP, generating the
C-terminal fragment of APP composed of 99 residues (CTFP or C99). CTFP is further cleaved by y-secretase to generate AP40/42. Alternatively, APP can be cleaved at the middle of the AP sequence by a.-secretase, resulting in soluble APPa. and membrane bound CTF a.. CTFa. is subsequently processed by y-secretase to release a small peptide with 24-26 residues called P3 that is not thought to be amyloidogenic [1]. Recently, we have identified a novel presenilin dependent cleavage site, the /;-site, within the TM domain of APP, generating Ap46 [139].
The subcellular sites forthe generation of AP have been the subject of intensive research. y-secretase that catalyzes cleavages in APP TM domain and determines the C-terminal heterogeneity of AP is known to consist of at least four proteins, presenilin (PS), 73 nicastrin, Aphl and Pen2. Presenilins (PSs) are believed to constitute the catalytic subunit of y-secretase([89, 94, 95, 97, 98] ). Althouth PS 1 is predominantly located in the endoplasmic reticulum (ER), the intermediate compartments, and the cis-Golgi compartment [172-175], the major y-secretase activity has been shown to be in the late compartments of the secretory pathway, at the cell surface, and in endosomes[l 76-179].
After being synthesized in the endoplasmic reticulum (ER), APP matures and is processed through the secretory pathway to the cell surface, where a portion of APP has been shown to be reintemalized into endosomes/lysosomes through the endocytic pathway. The early subcellular site for generating A� has been reported, in particularly in neuronal cells, to be the ER and intermediate compartment(IC), where intracellular A�42 was generated [167, 180, 181]. Whereas overwhelming evidence has shown that the major sites forgeneration of secreted A�40 and A�42are in the Golgi/TGN and secretory vesicles[167, 176, 182, 183]. A portion of A� has also been demonstrated to be generated on the cell surface [184], and in endosomes/lysosomes through reintemalization fromthe cell surface [177, 185]
Since the intracellular and secreted A� are knownto be generated at distinct subcellular sites [186, 187], little is known regarding the precise intracellular site in which A�46 is generated. To better understand . the role of �-cleavage and the relationship between � and y-site cleavages, it is necessary to determine the A�46 generation subcellular site.
Here we have examined the subcellular locations of A�46 generation using different approaches. We found that A�46 was presented in 20k xg crude membrane pellet but not in 20k xg supernatant, and the majority of A�46 was resistant to cell surface trypsinization. Treatment with BF A abolished the production of A�46, whereas monensin
74 treatment slightly affected AP46 level. Proteasomal inhibitor lactacystin, and endosomal/lysosomol inhibitor NH4Cl did not block AP46 generation. Furthermore, results from both transient transfection and stable clone showed that addition of an ER sorting signal (KKQN) to the C-terminus of APPsw completely blocked AP46 generation while addition of a TGN sorting signal sequence (SDYQRL) did not affect l;-cleavage.
Our data collectively suggest that AP46 is mainly generated in the Golgi/TGN compartments in N2a cells.
5.2 Materials and Methods
5.2.1 Plasmid construction. cDNA fragments encoding organelle sorting signals either
KKQN( designated as APPsw-ER) or SDYQRL( designated as APPsw-TGN) C terminally tagged APPsw695 were generated by PCR based mutagenesis usmg
APPswe695 cDNA as the tempelate. The cDNA fragments were ligated into Topo TA cloning vector (lnvitrogen) and subcloned into Hindlll and Xbiol sites of pcDNA3.1 vector (invitrogen) to get APPsw-ER/pcDNA3. l and APPsw-TGN/pcDAN3.1 plasmids, respectively. All cDNAs were verifiedby sequencing.
5.2.2 Cell lines. Mouse neuroblastoma cells lines stably expressing either human wild type PS1 (N2aPS1 wt) or both C-terminally myc tagged Swedish mutant APP and human wild-type PS 1 (N2a APPsw/PS 1 Wt) were established as described previously.
N2aAPPsw-ER/PS1 wt and N2aAPP-TGN/PS 1 wt cell lines stably expressing human wild-type pS 1 plus Swedish mutant APP695 C-terminally tagged either with ER sorting motif or with TGN sorting signal were established by transfection of N2aPS 1 wt cells with either APPsw-ER/pcDNA3.1 or APPsw-TGN/pcDNA3.1 vectors and subsequent selection of stable clones with G418 drug. 75 5.2.3 Antibodies and chemicals. Monoclonal antibodies 22C 11 and 6E 10 which recognizes residues 66-81 of APP and residues 1-17 of the A� sequence were purchased from Boehringer Mannheim and Senetek, respectively. Monoclonal antibodies anti-Bip and anti-y-adaptin were fromBD Transduction laboratories. Polyclonal antibody anti-Grp was from Affinity Bioreagents. Polyclonal antibody C 15 against the last 15 residues of
APP was described previously[139]. Rhodamine Red-X-Conjugated anti-mouse antibody and Alexa-Fluor 488 anti-rabbit antibody were from Jackson Immuno Research
Laboratories and Molecular Probes, respectively.
5.2.4 Preparation of crude membrane fraction. N2aAPPsw/PS 1 wt cells cultured in the presence of 0.5 µM DAPT for 4 h were homogenized in homogenization buffer( 20mM
HEPES, pH7.4, 50mM KCl, 2mM EGTA, 10% glycerol with protease inhibitor cocktail)
[142] by passing through a 20 gauge needle 30 times. The homogenate was centrifuged at 800 x g for 1 Omin to yield the 800 x g pellet (P800g) and supernatant (post nuclear supernatant). The resulting supernatant was further centrifuged at 20k x g for 20min to get the crude membrane fraction (P20kxg) and supernatant (S20kxg). Both P20kxg and
S20k xg fraction were dissolved in lysis buffer and equal amounts of protein . were subjected to westernblot.
5.2.5 Surface Trypsinization. N2aAPPsw/PS1 wt cells grown in three 10-cm dishes at
,_,75% confluence were treated with 0.5µM DAPT for 4 h to accumulate A�46, followed by washing with cold IX PBS twice. Cells were next incubated at 37°C with 6mL 1 X
PBS (0.5µM DAPT), 3mL lX PBS plus 3mL0.25% trypsin-EDTA solution (Sigma), or
6mL 0.25% trypsin-EDTA solution (Sigma), respectively. Both IX PBS and 0.25% trypsin-EDTA solution contained 0.5µM DAPT to prevent accumulated A�46 andCTF�
76 from turnover. After 30 min trypsinization, cells were collected in 20 mL fresh DMEM medium containing 10% FBS, centrifuged to get cell pellet, and further washed with
20mL fresh medium twice before adding 0.5mL Lysis-buffer (50mM Tris-HCl, pH6.8,
2%SDS, 8M urea, 2mMEDTA, 2mM DTT, and protease inhibitor cocktail ). After brief sonication, lysates were mixed with SDS-PAGE sample loading buffer and boiled at 100
°ಢC for 5.
5.2.6 Organelle toxin/inhibitor treatment. N2aAPPsw695/PS 1 wt cells that stably express Swedish mutant APP695 and wild-type PS 1 were grown in 10-cm dishes for24 h to reach semi-confluence in DMEM containing 10% FBS. The media were then replaced withfresh media containing either vehicle ( as a control) or one of the following inhibitors: 10 µM brefeldin A (BFA, 10 mM stock solution in DMSO, Sigma), 5 µM or
10 µM monensin (10 mM stock solution in ethanol , Sigma), 10 mM or 25 mM N"4Cl
(5M stock solution), 10 µM lactacystin (10 mM stock solution in DMSO, Calbiochem), or 10 µg/ml tunicamycin ( 10 mg/ml stock solution in DMSO, INC Biomedicals) as indicated in each figure. After culturing for 8 h, the conditioned media were collected, and the cells were scraped in Lysis Buffer (50mM Tris-HCl, pH6.8, 2%SDS, 8M urea,
2mMEDTA, 2mM DTT, and protease inhibitor cocktail(Roche) ) containing 0.5 µM
DAPT (Calbiochem) , briefly sonicated, mixed with sample loading buffer, and boiled for westernblotting
5.2.7 Immunoprecipitation and western blotting. Western blotting was performed essentially as described previously [139]. 7.5% SDS-PAGE was used for analysis of soluble APP fragments or nicastrin; 8-18% gel was used to analyze C-terminal fragments of APP or A� peptides; 10% Bicine/urea SDS-PAGE system was employed to separate
77 A� species as described previously[l 39].
5.2.8 Immunofluorescence microscopy. N2aAPPsw/PSlwt cells grown on microscope cover glasses (Fisher brand) in the presence or absence of 10 µM BFA or 5 µM monensin for Sh, or the cells that stably express C-terminally tagged APPsw (APP-ER and APP
TGF) grown on microscope cover glasses, were fixed with 4% paraformaldehyde at room temperature for 30 min, followed by washing with cold lX PBS three times. The fixed cells were permeabilized with 0.3%Triton X 100 for 5 min , washed with cold PBS three times, and blocked with blocking solution (5% lamb serum, 0.1 % Tween, lX PBS) at room temperature for 45 min. The cells were next incubated with appropriate primary antibodies (as indicated in each figure) in blocking solution at 4 °C overnightand washed with lX PBS containing 0.075% tween-20 five times. Cells were then incubated with either Rhodamine Red-X-Conjugated anti-mouse antibody (Jackson immuno Research laboratories) or Alexa-Fluor 488 anti-rabbit antibody (molecular probe) for 1 h followed by washing with lX PBS containing 0.075% tween -20. The cells were visualized under
Niko Eclipse E600 microscope and immunofluorescence images were captured using
Qcapture software.
5.3 Results
5.3.1 AP46 is present in the crude membrane fraction. Having shown that major A�46 is detected in total cell lysates but not in conditioned medium [139], prompted us to determine the intracellular compartments of A�46 generation. We first performed a brief centrifugation to separate cytosolic proteins from the large membrane. As shown in
FiglA, CTF� was detected in both supernatant and pellet, consistent with previous report that ·�-secretase shows highest activity in various compartments including Golgi, TGN, 78 and secretory vesicles[72]. In contrast, A�46 was only detected in P20k xg fraction, suggesting that A�46 is predominantly located in large membrane compartments such as the ER, the Golgi/TGN, or the plasma membrane. Similar results were reproduced from compound E treated cells ( data not shown), indicating that y-secretase inhibitors do not alter the localization of A�46. Interestingly, both PS 1 and Nicastrin were also mainly detected in 20k xg pellet fraction (data not shown), indicating that A�46 may be co localized with the y- secretase complex.
5.3.2 The majority of AP46 is resistant to surface trypsinization. We next determined whether Af346 resides on the cell surface by employing cell surface trypsinization as described by other investigators [184, 187]. Both A�46 and CTF� contain three putative cleavage sites for trypsin (the region outside of the plasma membrane): 5Arg , 16Lys, and 28 Lys of the A� sequence. Cleavage at 5Arg would result in a small A� or a CTF fragment, whereas cleavages at 16Lys or 28Lys destroythe recoganization for 6E10. We found that surface trypsinization dramatically decreased the CTF� level (Fig2, lanes2-3), especially at high doses of trypsin (Fig2, lane 3). This suggests that CTF� accumulated by treatment with DAPT is mainly localized on the cell surface, or it may be derived from cell surface localized APP during the last 30 min before lysis. In contrast, the A�46 level was only slightly reduced by cell surface trypsinization, even at high dose of trypsin
(Fig2, lanes 2 and 3). This finding indicates that A�46 is mainly present in intracellular compartments, and only a small portion of A�46 exists on the cell surface.
79 5.3.3 Treatments with BFA and monensin block APP secretion in N2a cells, whereas lactacystin, ammonium chloride, and tunicamycin do not inhibit APP secretion.
Although the brief centrifugation and surface trypsinizaiton gave us a clue of the localization of A�46, the place in which A�46 exists may not truly reflectthe subcellular compartments in which the A�46 is generated, thus we employed protein trafficking inhibitors to investigate the sites of A�46 generation. Both BFA andmonensin have been widely used to study protein trafficking pathway. The fungal metabolite BF A has been shown to block anterograde transport and maturation of APP between the ER and the
Golgi[187-1 89], causing the proximal Golgi to fuse with the ER, and the TGN to fuse with endosomes. Ionophore monensin is known to block protein trafficking and maturation within the GolgiffGN, and neutralizes some acidic intracellular compartments such as the TGN, lysosomes, and certain endosomes[190, 191]. To investigate whether these inhibitors cause retention of over-expressed Swedish mutant APP in certain compartments in N2aAPPsw/PS 1 wt cells, we first cultured cells in the presence or absence of these inhibitors for 5 h, followed by double immunofluorescence staining of the cells with anti-APP antibody (6E10 or C-15) plus either anti-Grp/Bip antibody or anti-y-adaptin antibody to examine the localization of APPsw protein in N2a cells.
Grp78/Bip (glucose-regulated protein 78 or immunoglobulin binding protein) is a well characterized ER chaperone[192, 193], whereas y-adaptin is a TGN resident protein[l94].
As shown in Fig3, over-expressed APPsw in untreated N2a cells did not co-localize with
Grp, but a portion of APPsw did co-localize with y-adaptin, which is in agreement with previous reports showing that APP is concentrated in the Golgi complex [195]. As expected, treatment with BF A led to co-localization of APPsw with Grp/Bip, whereas
80 treatment with monensm resulted in the co-localization of APPsw with y-adaptin, indicating that BF A and monensin treatments retain over-expressed APPsw predominantly in the ER and TGN in N2a cells, respectively, which is consistent with other's observations[l 95].
We next examined the effects of BF A, monensin, lactacystin, ammonium chloride and tunicamycin on APP secretion. N2aAPPsw/PS1 wt cells were treated with these inhibitors at indicated concentrations for 8 h. Then both conditioned media and cell lysates were analyzed by western blot. Consistent with previous reports[195],treatment with BFA led to elevated levels of intracellular full-length APP (Fig3A. lane 2 in all three panels) and completely abolished APP secretion (Fig4B, lane 2 in both panels), indicating that APP processing occurs after it leaves the ER. Interestingly, two clear bands were observed from monensin treated cell lysate when probed with 6E10 antibody (Fig4A top panel, lanes 3-4), whereas only the top band was shown when probed either with C15 antibody or with anti-myc antibody (Fig4A, medium and bottom panels, lanes 3-4). As myc is tagged at the C-terminus of APPsw, and C15 antibody recognizes the C-terminal 15 residues of APP, the band thatonly can be detected by 6E10 (Fig4A, top panel, lanes3-4, lower band), but not by C15 and anti-myc antibodies is most likely the N-terminus of
APP cleaved at the a-site (APPa), because �-secretase generated N-terminal fragment of
APP, APP�, can not be recognized by 6E10, whereas y-secretase mediated cleavages are believed to occur after �- or a-cleaveges. Although the total secreted APP ( detected with
22C 11 antibody which recognizes residues 66-81 of APP) was reduced in the conditioned media from monensin treated cells (Fig4B, compare lanes3-4 with lane 1), APPa
( detected with 6E10 which recognizes residues 1-17 of A� sequence) secretion was
81 slightly increased as compared with that from untreated cells (Fig4B, bottom panel, compare lanes 3,4 with lane 1). The increase of secreted APPa may not be due to the failure of monensin to block APP secretion, but because of the markedly elevated intracellular level of APPa (Fig4A, top panel, lanes3, 4). Retention of APPsw within the
TGN by treatment with monensin elevated level of full-length APP, the substrate of a secretase, as a result of the inhibition of APP nonspecific degradation (Fig4 A, lanes3,4 in all three panels),thus leading to increased level of a-secretase cleavage product, AP Pa.
This finding also confirms previous observations that the TGN is one of the major subcellular compartments for a-secretase mediated APP processing [70, 196, 197].
Treatment with monensin slowed protein secretion but did not completely abolish protein trafficking, which is consistent with previous observations[190, 191]. Inhibition of the proteasome with lactacystin[l 98] did not alter APP secretion (Fig4B, lane5 in both panel), however, lactacystin slightly decreased full-lengthAPP level (Fig4A, compare lane5 with lane 1 in the medium and bottom panels). The reason why the inhibition of the proteasomal degradation pathway affects exogenousAPP expression is currently unclear.
However inhibition of the proteasomal degradation pathway may result in several effects such as apoptosis, and differentiation of neuroblastoma cells[199].
Since the endosomal/lysosomal pathway has also been shown to be involved in APP processing [185, 200], the effect of NH4Cl on APP secretion was also examined. As shown in Fig4B (lanes 6, 7 in both panels), NH4Cl treatment did not block, but rather slightly increase APP secretion. Interestingly, when we examined intracellular APP level with 6E10 antibody, we foundtwo bands fromNH4Cl treated samples (Fig4A, top panel, lanes 6,7). However the lower band was not recognized by C15 or anti-myc antibodies,
82 suggesting that it is most likely APPa (Fig4A, medium and bottom panels, lanes6,7). The similar patterns obtained from monensin and NH4Cl treatments suggest that their effects overlap to some extent. However, the lower band in Fig4A, top panel, lanes 6, 7 shows lower mobility than the lower band in lanes3,4 of Fig4A, top panel, which possibly results from different post-translational modifications of APPa. Monensin treatment partially blocks APP post-translational modification of APP [191], thus leading to lower molecular weight APPa, whereas NH4Cl treatment mainly affects endosomal/lysosomal functions, and would be less likely to affect post-translational modification. This finding also suggests that immature APP can be processed by a-secretase.
Since APP contains two potential N-glycosylation sites at Asn467 and Asn 496, we also incubated cells with tunicamycin, which blocks cotranslational N-glycosylation within the ER[201]. However, inhibition of APP secretion was only observed when detected with 6E10(Fig4B,bottom panel, lane8), but not with 22Cll (Fig4B top panel, lane8), suggesting that inhibition of N-glycosylation of APP only affects a-secretase mediated processing but not �-site cleavage.
5.3.4 BFA treatment abolishes AP46 generation, monensin only slightly reduces
AP46 level, whereas ammonium chloride, lactacystin and tunicamycin do not affect
AP46 generation. To investigate the subcellular locations of A�46 generation, we treated cells with the above described trafficking or organelle inhibitors in the presence or absence of compound E, then analyzed their effects on A�46 generation. We employed both regular 10/18% SOS-PAGE and 10% Bicine/urea SOS-PAGE systems to confirm the identities of intracellular A�, and standard A�40/42markers were also included. As expected, treatment with compound E led to intracellular accumulation of A�46 (Fig5A,
83 lane 9 in both panels). Surprisingly, treatment with BF A completely abolished A�46 generation (Fig5A, lanelO in both panels). Although A�42 has been reported to be generated in the ER in several cell lines[l 67, 180, 181 ], A�42 generated in the presence of BF A was undetectable by directly loading lysate without immunoprecipitation (Fig5A, top panel, lane 10), demonstrating that both A�46 and A�42/40 are less likely generated in the ER from full-lengthAPP in N2a cells.
We next examined whether retention of APP at the Oolig/TON by monensin affects
A�46 production. As shown in Fig5A lanesll and 12, in both panels, 5µM monensin largely did not affect A�46 level, whereas 10µM monensin reduced A�46 produciton.
Since monensin is known to inhibit TON function [191], these findings suggest that
A�46 can be generated in Oolgi and TON subcellular sites. There are at least two possibilities for the reduction of A�46 at high concentration of monensin: 1) A�46 is predominantaly generated in the Oolgi and TON, but high concentration of monensin affects the generation of A�46 in TON as a result of inhibition of TON function by monensin; 2) a portion of A�46 may be generated downstream of the TON. It is noted that the CTFP level was dramatically decreased in monensin treated cells, especially at high concentration of monensin (Fig5A, lanes 11, 12 in both panels). Thus the reduction of A�46 generation at high concentration of monensin is very likely due to the inhibition of �-secretase mediated cleavage, rather than �-cleavage.
We also examined the effect of lactacystin on A�46 production, as shown in Fig5A lane 13 in both panels, although CTF� level was decreased, inhibition of the proteasome withlactacystin did not reduce A�46 level, indicating that the proteasome is not involved in generation of A�46. To examine whether endosomes/lysosomes are involved in A�46
84 production, we also treated cells with both NH4Cl and compound E. As shown in Fig4A lanes 14 and 15 in both panels, 25mM NH4Cl did not inhibit A�46 production, indicating that endosomes/lysosomes are not involved in A�46 generation. Tunicamycin did not affect A�46 level (Fig5A, lane 16 in both panels), suggesting that N-glycosylation of
APP is not required for generation of A�46.
The effects of theseinhibitors on the level of AICD were also determined. No C-termini of APP were observed in BFA treated cells even cultured in the presence of y- secretase inhibitor compound E (Fig5B,both panels, lane 10), indicating the absence of any secretase-mediated processing in the ER. In contrast to BF A, treatment with monensin led to elevated level of both endogenous and exogenous AICD (Fig5B, lanes3,4,6 and 7), demonstrating that e-cleavage of APP occurs before APP exits from the TGN. A conceivable reason for the increase of AICD in monensin treated cells is that monensin treatment prevents AICDfrom leaving the Golgi/TGN fordownstream degrada tion, since
AICD is known for its very short half life. As expected, inhibition of endosome/lysosome functionwith NH4Cl dramatically enhanced accumulation of AICD, confirming that AICD is unstable, and the major site for degradation of AICD is endosomes/lysosomes. Consistent with previous finding,y-secretase inhibitor compound
E partially blocked e-cleavage, thus diminished AICD generation (Fig5B compare lanesl 1,12,14, 15 with lanes3,4,6,7)
5.3.5 AP46 and Ap40/42are differentially affected by monensin, NH4Cl, lactacystin, and tunicamycin. We then examinedthe effects of these inhibitors on A�40/42secretion.
Similarly to the effect of BFA on A�46, we did not find any A� species in the conditioned medium(Fi g5C, lane3) or the lysate (Fig5A, lane3 in both panels) fromBFA
85 treated cells. Although A�42 has been reported to be generated in the ER [167, 181, 202], we and others did not confirm these findings[l 72, 203-205]. Although we can not rule out the possibility that a trace amount of A�42 may exist in the lysate of BFA treated cells, as we directly loaded the lyaste without immunopreciation, the amount of intracellular A�42 must be very low, since the endogenous level of A�46 was seen under the same condition. Thus our result demonstrates that the ER is not a major site for generation of A�40 or A�42 (Fig5A, lane1 in both panels). Interestingly, 5 µM monensin which displayed almost no inhibitory effect on A�46 level (Fig5A, lane 11 ), clearly reduced A�40/42production (Fig5C, lanes4 and 7), whereas 10 µM monensin almost completely blocked both A�40 and A�42 secretion (Fig5C, lane4), indicating that the subcellular site for generation of A�46 is different from that for A�40/42. In addition,
NH4Cl treatment, which exhibited no effect on A�46 production under the test condition, led to a dose dependent inhibition of both A�42 and A�40 generation, further indicating that the site for A�46 generation is at earlier stage in the secretory pathway than that of
A�40/42. Furthermore, inhibition of proteasomes with lactacystin, and inhibition of N glycosylation with tunicamycin all blocked A�40/42generation (Fig5C, lanes 6 and 9), while these drugs exhibited no effect on A�46 generation (Fig5A, lanes 13 and 16), suggesting that A�46 and A�40/42are generated at distinct subcellular sites or by different enzymes. The parallel inhibition of both A�40 and A�42 for all the tested inhibitors suggests that A�40/42may be generated at the same compartments in this cell line.
86 5.3.6 Retention of APP in the ER by ER sorting motif diminishes both AP46 and
AP40/42generations, whereas addition of TGN sorting sequence to the C-terminus of
APP does not affect AP46 or Ap40/42generation. It is noted thattrafficking inhibitors
BF A and monensin block all protein transport and maturation, and monensin treatment leads to the inhibition of acidification of the TGN and lysosomes. To exclude the possibility that the inhibition of A�46 generation by BFA is due to the blockage of the maturationof other components that arerequired for A�46 generation, and under normal conditions, these components can reenter the ER after maturation, we generated an expression vector encoding Swedish mutant APPsw C-terminally tagged with ER sorting signal KK.QN(designated as APPsw-ER), which has been known to effectively retain protein in the ER[180, 204, 206, 207]. To confirm the major site for A�46 generation is the Golgi/fGN, we constructed an expression vector encoding Swedish mutant APP C-terminally tagged with TGN sorting signal SDYQRL(designated as
APPsw-TGN), which is known forretention of protein in the TGN [208, 209]. N2aPSlwt cells stably expressing human wild-type PS 1 were transiently transfected with either
APPsw-ER, APPsw-TGN, APPsw, or Lacz ( all in pcDNA3.1 plasmids), respectively, and cultured in thepresence or absence of 3nM compoundE or O.SµM L685,458 for 12h, followedby analysis of APP processing by westernblotting.
We first examined the expression and secretion of APP in transiently transfected cells.
Consistent with the traffickinginhibition results and findingsof other using the same ER sorting signal[204], retention of APP within the ER diminished APP secretion
(Fig6A,top panel, lanes 4-6), whereas retention of APP in TGN did not affect APP secretion but rather slightly increased secretion as compared with APPsw vector( Fig6A,
87 top panel, lanes7-12). When we determined the APP level in lysates, we found a concomitant increase of full-length APP in APPsw-ER transfected cells (Fig6A bottom panel lanes 4-6) and a concomitant decrease in APPsw-TGN transfected cells (Fig6A, bottom panel, lanes7-9). Due to the less homology between mouse APP and human APP within the 1-1 7 residues of the A� sequence, we did not detected the mouse endogenous
APP in Lacz transfected cells (Fig6A, lanesl-3 in both panels) when probed with 6E10.
We next determined the effect of ER or TGN sorting signals on A�46 generation. No
CTF-� or A�46 band was detected in lacZ transfected cells when probed with 6E10
(Fig6B, lanesl-3). Although a small portion of CTF�-ER was seen (Fig6B lanes4-6), only a very faint A�46 band was observed in APPsw-ER transfected and compound E treated cells (Fig6B lanes4-6),indicating that A�46 is not generated in the ER. Consistent with the result of monensin treatment, addition of TGN signal did not affect A�46 level
(Fig6B, compare lane8 with lane10), confirming that the major location of A�46 generation is within or upstream of the TGN.
We also examined the secreted A� level in these different vector transfected cells. We found that addition of TGN sorting signal did not affect total A� production (FigSC, compare lane7 with lanel 1) or A�42 level (Fig6D, compare lane 3 with lane4), while retention of APP in the ER significantly diminished the generation of both total
A�( Fig6C, lane4) and A�42(Fig6D, lane2), which is also consistent with previous report
[204], indicating that both A�40 and A�42 are generated downstream of the ER, but upstream of or within the TGN.
To confirm the transient transfection results, stable clones expressing either APPsw-ER or APPsw-TGN proteins were established by transfectionof N2aPS 1 wt cells either with
88 APPsw-ER or with APPsw-TGN plasmids, using G418 to select stable clones. Although the amount of full-lengthAPP in APPsw-ER clones was higher thanthat in APPsw-TGN clones (Fig6E, top panel,compare lanes l-6 with lanes 7-12), neither AP46(Fig6E, lanes2 and 5, treated with compound E) nor AP40/42( Fig6F, lanes land 2) was observed in
APPsw-ER clones, further confirming previous findings that retention of APP in the ER diminishes AP generation and APP secretion[204, 210]. Both AP46 (Fig6E, lanes 8 and
11, with compound E) and AP40/42 were clearly seen in the samples fromAPPsw-TGN clones. This finding further confirms that retention of APP with an ER sorting signal abolishes both AP46 and AP40/42generation, whereas addition of a TGN sorting signal does not affecteither AP46 or AP40/42production.
5.3.7 APPsw-ER and APPsw-TGN reside in the ER and the TGN in N2a cells, respectively. To ascertain the localizations of APPsw-ER and APPsw-TGN proteins, we performed double immunofluorescence microscopy using stable clones of APPsw-ER and APPsw-TGN. As expected, APPsw-ER was co-localized with Grp78 protein, an ER resident protein (Fig7), whereas APPsw-TGN was co-localized with y-adaptin (Fig7), a
TGN marker. These results confirm that APPsw-ER and APPsw-TGN locate in the ER and the TGN in N2a cells, respectively.
5.4 Discussion
Although various subcellular compartments including the ER, the IC, the Golgi/TGN, endosomes/lysosomes, and the cell surface have been demonstrated to be sites for AP generation [167, 176, 177, 180, 182-185], it has been known thatdifferent AP species are generated in distinct subcellular compartments, and the sites for generation of intracellular cellular and secreted AP are also different [167, 180, 181, 182]. Previously 89 we found a new A� species, sA� or A�46, which predominately exists in cell lysate but not in conditioned medium. To better understandthe mechanism of y-secretase mediated cleavages, we have determined that Golgi/TGN is the major site for A�46 generation.
We first determined whether A�46 resides in large membranes, vesicles, microsomes, or cytosol by centrifugation of post nuclear supernatant at 20k xg. Although CTF� was foundin bothsupernatant and pellet, A�46 was only seen in 20k xg pellet, indicating that
A�46 mainly exists in large membranes such as the ER/IC, the Golgi/TGN, or the cell surface, but less likely in cytosol or microsomes. To examine whether A�46 resides on the plasma membrane, we performed surface trypsinizaiton. Unlike CTF�, which was sensitive to surface trypsinization, the majority of A�46 was resistant to surface trypsinizaiton, indicating only a small amount of A�46 is located on the cell surface.
Since the localization of A�46 may not truly reflect the place where it is generated, we employed BF A and monensin, two known protein trafficking inhibitors, to examine the subcellular compartment for A�46 generation. Using double immunofluorescence staining, we confirm that APP was co-localized with Bip or y-adaptin, after treatment with BFA or monensin, respectively. Consistent with previous observations[l 95], treatment with BFA diminished the APP secretion, and elevated the level of full-length
APP. In contrast, treatment with monensin reduced total soluble APP (sAPP� and sAPPa.) level, but led to slightly elevated level of soluble APPa and accumulation of large amounts of intracellular APPa fragment. Treatment with NH4Cl increased APP secretion and intracellular accumulation of APPa, suggesting that NH4Cl inhibits the degradation of APP in endosomes/lysosomes. We next examined A�46 level in various inhibitor treated cells. Surprisingly, we found treatment with BFA completely abrogated A�46
90 generation even in the presence of compoundE, indicating that A�46 can not be produced in the ER. Monensin only exhibited inhibitory effect on A�46 generation at high concentrations, suggesting that the site for Abta46 generation is in the Golgi and the TGN.
The partial reduction of A�46 generation at high dose of monensin was likely caused by the blockage of A�46 generation in the TGN as a result of the inhibhition of TGN function by monensin. The finding that the CTF� level from the cells treated with a high dose of monensin was dramatically reduced (Fig5, lane 12 in panels A and B) favorsthis suggestion. Moreover, the fact that neither inhibition of endosomes/lysosomes with
NH4Cl nor inhibition of proteasome with lactacystin affected A�46 level, demonstrates that A�46 is not generated in these organelles. Inhibition of N-glycosylation of APP by tunicamycindid not reduce A�46 level, indicating that N-glycosylation is not required for
A�46 generation.
To furtherconfirm that A�46 is generated mainly in the Golgi/TGN but not in the ER, we added ER or TGN sorting motif to the C-terminus of Swedish mutant APP. Transient transfection results show that addition of ER sorting motif diminished both APP secretion and A�46 generation, whereas addition of TGN sorting signal did not affect A�46 level, nor APP secretion, as compared with APPsw. APPsw-ER and APPsw-TGN stable clone results also show that A�46 was only seen in APPsw-TGN cells but not in APPsw-ER cells. Furthermore, co-localization of APPsw-ER with Bip/Grp and co-localization of
APPsw-TGN with y-adaptin were also observed by double immunofluoresence staining.
These findings demonstrate that retention of APP in the ER abolished A�46 generation whereas retention of APP in the TGN did not affect A�46 level. Thus A�46 is predominantly generated in the Golgi/TGN in this cell line.
91 The finding that A�46 is generated in the Golgi/TGN is consistent with the previous observations showing that �-secretase activity has been localized in the Golgi/TGN[72], whereas assembled y-secretase complex has been reported to localize predominantly in the Golgi and to a lesser extent, at the cell surface and in endocytic compartments [175,
210-212]
Another interesting finding of the present study is that s-cleavage and y-cleavage are differentially affected by these inhibitors. As shown in Fig5, monenisn treatment at 5µm did not affectA�46 level but clearlyreduced A�40/42level, indicating that A�46 may be generated at the earlier stage in the secretory pathway than A�40/42. The e-cleavage product AICD was still seen when cells were treated with 10µM monensin, whereas
A�40/42level was almost completely blocked, suggesting e-cleavage of APP also likely occurs at the earlier stage than y-secretase cleavage of APP. NH4Cl treatment results further support the above suggestion, since 25mM NH 4Cl did not exhibited any inhibitory effect on A�46 and AICD generation, while both A�40 and A�42 levels were dramatically diminished by NH4Cl. These observations indicate that both E- and s cleavages of APP occur at earlier stage of the secretory pathway than that of y-cleavage of APP. Lactacystin and tunicamycin treatments exhibited similar results on
A�40/42level but did not affect A�46 generation, indicating that A�46 and A�40/42are generated at distinct subcellularsites or generated by similar but differentenzymes.
Taken together, our data indicate that A�46 is generated at the Golgi and TGN compartments, whereas A�40 and A�42 are likely generated at the TGN and downstream compartments. This finding may provide insight for explaining the discrepancy between the localization of PSs and the A� generation sites.
92 REFERENCES
93 References:
1. Selkoe, D.J., Alzheimer's disease: genes, proteins, and therapy. Physiol Rev, 2001. 81(2): p. 741-66. 2. Selkoe, D.J. and D. Schenk, Alzheimer's disease: molecular understanding predicts amyloid-based therapeutics. Annu Rev Pharmacol Toxicol, 2003. 43: p. 545-84. 3. Mandelkow, E., Alzheimer's disease. The tangled tale of tau. Nature, 1999. 402(6762): p. 588-9. 4. Grundke-lqbal, I., et al., Abnormal phosp horylation of the microtubule-associated protein ta u (tau) in Alzheimer cytoskeletal pathology. Proc Natl Acad Sci U S A, 1986. 83(13): p. 4913-7. 5. von Bergen, M., et al., Ta u aggregation is driven by a transition fromrandom coil to beta sheet structure. Biochim Biophys Acta, 2005. 1739(2-3): p. 158-66. 6. Drewes, G., MARKingta ufor tangles and toxicity. Trends Biochem Sci, 2004. 29(10): p. 548-55. 7. Dickson, D.W., Th e pathogenesis of senile plaques. J Neuropathol Exp Neurol, 1997. 56(4): p. 321-39. 8. Glenner, G.G. and C.W. Wong, Alzheimer's disease and Down 's syndrome: sharing of a unique cerebrovascular amyloidfibril protein. Biochem Biophys Res Commun, 1984. 122(3): p. 1131-5. 9. Glenner, G.G. and C.W. Wong, Alzheimer's disease: initial report of the purificationan d characterization of a novel cerebrovascular amyloid protein. Biochem Biophys Res Commun, 1984. 120(3): p. 885-90. 10. Jarrett,J.T., E.P. Berger, and P.T. Lansbury, Jr., Th e carboxyterminus of the beta amyloid protein is critical fo r the seeding of amyloidfo rmation: implications fo r the pathogenesis of Alzheimer's disease. Biochemistry, 1993. 32(18): p. 4693-7. 11. lwatsubo, T., et al., Visualization of A beta 42(43) and A beta 40 in senile plaques with end-specificA beta monoclonals: evidence that an initially deposited species is A beta 42(43). Neuron, 1994. 13(1): p. 45-53. 12. Selkoe, D.J., The molecular pathologyof Alzheimer's disease. Neuron, 1991. 6(4): p. 487-98. 13. Hardy,J.A. and G.A. Higgins, Alzheimer's disease: the amyloid cascade hyp othesis. Science, 1992. 256(5054): p. 184-5. 14. Goldgaber, D., et al., Characterization and chromosomal localiza tion of a cDNA encoding brain amyloid of Alzheimer's disease. Science, 1987. 235(4791): p. 877- 80. 15. Kang, J., et al., The precursor of Alzheimer's disease amyloid A4 protein resembles a cell-surfa ce receptor. Nature, 1987. 325(6106): p. 733-6. 16. Tanzi, R.E., et al., Amyloid beta protein gene: cDNA, mRNA distribution, and genetic linkage near the Alzheimer locus. Science, 1987. 235(4791): p. 880-4. 17. Olson, M.I. and C.M. Shaw, Presenile de mentia and Alzheimer's disease in mongolism. Brain, 1969. 92(1): p. 147-56.
94 18. Hardy,J. and D.J. Selkoe, The amyloid hyp othesis of Alzheimer's disease: progress and problems on the road to therapeutics. Science, 2002. 297(5580): p. 353-6. 19. Cai, X.D., T.E. Golde, andS.G. Younkin, Release of excess amyloid beta protein from a mutant amyloid beta protein precursor. Science, 1993. 259(5094): p. 514- 6. 20. Citron,M., et al., Mutation of the beta-amyloid precursor protein in fa milial Alzheimer's disease increases beta-protein production. Nature, 1992. 360(6405): p. 672-4. 21. Suzuki,N., et al., An increased percentage of long amyloid beta protein secreted by fa milial amyloid beta protein precursor (beta APP7 1 7) mutants. Science, 1994. 264(5163): p. 1336-40. 22. Sherrington, R., et al., Cloning of a gene bearing missense mutations in early onset fa milial Alzheimer's disease. Nature, 1995. 375(6534): p. 754-60. 23. Scheuner, D., et al., Secreted amyloid beta-protein similar to that in the senile plaques of Alzheimer'sdi sease is increased in vivo by the presenilin 1 and 2 and APP mutations linked to fa milial Alzheimer's disease. Nat Med, 1996. 2(8): p. 864-70. 24. Hutton, M., et al., Association of missense and 5'-splice-site mutations in ta u with the inherited dementia FTDP-17. Nature, 1998. 393(6686): p. 702-5. 25. Lewis, J., et al., Enhanced neurofibrillarydege neration in transgenic mice expressing mutant tau and APP. Science, 2001. 293(5534): p. 1487-91. 26. Gotz, J., et al., Formation of neurofibrillarytangles in P301l ta u transgenic mice induced by Abeta 42.fibrils. Science, 2001. 293(5534): p. 1491-5. 27. Bales, K.R., et al., Lack of apolipoprotein E dramatically reducesam yloid beta peptide deposition. Nat Genet, 1997. 17(3): p. 263-4. 28. Myers, A., et al., Susceptibilitylocus fo r Alzheimer's disease on chromosome 10. Science, 2000. 290(5500): p. 2304-5. 29. Kuo, Y.M., et al., Wa ter-soluble Abeta (N- 40, N- 42) oligomers in normal and Alzheimer disease brains. J Biol Chem, 1996. 271(8): p. 4077-81. 30. McLean, C.A., et al., Soluble pool of Abeta amyloid as a determinant of severity of neurodegeneration in Alzheimer's disease. AnnNeurol, 1999. 46(6): p. 860-6. 31. Naslund, J., et al., Correlation between elevated levels of amyloid beta-peptide in the brain and cognitive decline. Jama, 2000. 283(12): p. 1571-7. 32. Wirths, 0., G. Multhaup, andT.A. Bayer, A modifiedbet a-amyloid hyp othesis: intraneuronal accumulation of the beta-amyloid peptide--the firststep of a fa tal cascade. J Neurochem, 2004. 91(3): p. 513-20. 33. Cumming, J.N., U. Iserloh, andM.E. Kennedy, Designan d development of BA CE-1 inhibitors. CurrOpin Drug Discov Devel, 2004. 7(4): p. 536-56. 34. Rocchi, A., et al., Ca usative and susceptibilitygenes fo r Alzheimer's disease: a review. Brain Res Bull, 2003. 61(1): p. 1-24. 3 5. Goate, A., et al., Segregation of a missense mutation in the amyloid precursor protein gene with fa milial Alzheimer's disease. Nature, 1991. 349(6311): p. 704-6. 36. Mullan, M., et al., A pathogenic mutation fo r probable Alzheimer's disease in the APP gene at the N-terminus of beta-amyloid Nat Genet, 1992. 1(5): p. 345-7.
95 37. Mann, D.M., Alzheimer's disease and Down's syndrome. Histopathology, 1988. 13(2): p. 125-37. 38. Wisniewski, K.E., et al., Alzheimer's disease in Down's syndrome: clinicopathologic studies. Neurology, 1985. 35(7): p. 957-61. 39. Selkoe, D.J., Alzheimer's disease: genotypes, phenotypes, and treatments. Science, 1997. 275(5300): p. 630-1. 40. Bertram, L. and R.E. Tanzi, The current status of Alzheimer's disease genetics: what do we tell the patients? PharmacolRes, 2004. 50( 4): p. 385-96. 41. Borchelt, D.R., et al., Familial Alzheimer's disease-linked presenilin 1 variants elevate Abetal-42/1-40 ratio in vitro and in vivo. Neuron, 1996. 17(5): p. 1005-13. 42. Citron, M., et al., Mutant presenilins of Alzheimer's disease increase production of 42-residue amyloid beta-protein in both transfectedce lls and transgenic mice. Nat Med, 1997. 3(1): p. 67-72. 43. Duff, K., et al., In creased amyloid-beta42(43) in brains of mice expressing mutant presenilin 1. Nature, 1996. 383(6602): p. 710-3. 44. Strittmatter, W.J., et al., Apolipoprotein E: high-aviditybind ing to beta-amyloid and increasedfreque ncy of typ e 4 allele in late-onset fa milial Alzheimer disease. Proc Natl Acad Sci USA, 1993. 90(5): p. 1977-81. 45. Corder, E.H., et al., Gene dose of apolipoprotein E typ e 4 allele and the risk of Alzheimer's disease in late onset fa milies. Science, 1993. 261(5123): p. 921-3. 46. Saunders, A.M., et al., Association of ap olipoprotein E allele epsilon 4 with late onset fa milial and sporadic Alzheimer's disease. Neurology, 1993. 43(8): p. 1467- 72. 47. Corder, E_.H., et al., Protective effe ct of ap olipoprotein E typ e 2 allele fo r late onset Alzheimer disease. Nat Genet, 1994. 7(2): p. 180-4. 48. Weisgraber, K.H., Apolipoprotein E: structure-function relationships. Adv Protein Chem, 1994. 45: p. 249-302. 49. Mahley, R.W., Apolipoprotein E: cholesterol transport protein with expanding role in cell biology. Science, 1988. 240( 4852): p. 622-30. 50. Zannis, V.I., P.W. Just, and J.L. Breslow, Hu man ap olip oprotein E isoprotein subclasses are genetically de termined. Am J Hum Genet, 1981. 33(1): p. 11-24. 51. Olichney, J.M., et al., Th e ap olipoprotein E epsilon 4 allele is associated with increased neuritic plaques and cerebral amyloid angiopathy in Alzheimer's disease and Lewy bodyvariant. Neurology,1996. 47(1): p. 190-6. 52. Strittmatter,W.J., et al., Binding of human ap olipoprotein E to synthetic amyloid beta peptide: isoform-specific effe cts and implications fo r late-onset Alzheimer disease. Proc Natl Acad Sci US A, 1993. 90(17): p. 8098-102. 53. Sanan, D.A., et al., Apolipoprotein E associates with beta amyloid peptide of Alzheimer's disease to fo rm novel monofibrils. Js oform ap oE4 associates more effi ciently than ap oE3. J Clin Invest, 1994. 94(2): p. 860-9. 54. Strittmatter, W.J., et al., Hypothesis: microtubule instabilityand paired helical filamentfo rmation in the Alzheimer disease brain are related to ap olipoprotein E genotype. Exp Neurol, 1994. 125(2): p. 163-71; discussion 172-4. 55. Qiu, Z., et al., Alpha2-macroglobulin enhances the clearance of endogenous soluble beta-amyloid peptide via low-densitylipo protein receptor-related protein 96 in cortical neurons. J Neurochem, 1999. 73(4): p. 1393-8. 56. Lendon, C.L., et al., Genetic association studies between de mentia of the Alzheimer's typ e and three receptors fo r apolipoprotein E in a Caucasian population. Neurosci Lett, 1997. 222(3): p. 187-90. 57. Bertram, L., et al., Evidence fo r genetic linkage of Alzheimer's disease to chromosome JOq. Science, 2000. 290(5500): p. 2302-3. 58. Kehoe, P.G., et al., Variation in DCP 1, encoding ACE, is associated with susceptibilityto Alzheimer disease. Nat Genet, 1999. 21(1): p. 71-2. 59. Seubert, P ., et al., Secretion of beta-amyloid precursor protein cleaved at the amino terminus of the beta-amyloidpeptide. Nature, 1993. 361(6409): p. 260-3. 60. Esch, F .S., et al., Cleavage of amyloid beta peptide during constitutive processing of its precursor. Science, 1990. 248(4959): p. 1122-4. 61. Sisodia, S.S., et al., Evidence that beta-amyloid protein in Alzheimer's disease is not derived by normal processing. Science, 1990. 248(4954): p. 492-5. 62. Buxbaum, J.D ., et al.,Evid ence that tumor necrosis fa ctor alpha converting enzyme is involved in regulated alpha-secretase cleavage of the Alzheimer amyloidproteinprecursor. J Biol Chem, 1998. 273(43): p. 27765-7. 63. Lammich, S., et al., Constitutive and regulated alpha-secretase cleavage of Alzheimer's amyloid precursor protein by a disintegrin metalloprotease. Proc Natl Acad Sci USA, 1999. 96(7): p. 3922-7. 64. Koike, H., et al., Membrane-anchored metalloprotease MD C9 has an alpha secretase activityres ponsible fo r processing the amyloid precursor protein. Biochem J, 1999. 343 Pt 2: p. 371-5. 65. Hooper, N.M. and A.J. Turner, The search/or alpha-secretase and its potential as a therapeutic approach to Alzheimer s disease. Curr Med Chem, 2002. 9(11): p. 1107-19. 66. Lichtenthaler, S.F. and C. Haass, Amyloid at the cutting edge: activation of alpha secretase prevents amyloidogenesis in an Alzheimer disease mouse model. J Clin Invest, 2004. 113(10): p. 1384-7. 67. Etcheberrigaray,R., et al., Therapeutic effects of PKC activators in Alzheimer's disease transgenic mice. Proc Natl Acad Sci USA, 2004. 101(30): p. 11141-6. 68. LeBlanc, A.C., M. Koutroumanis, and C.G. Goodyer, Protein kinase C activation increases release of secreted amyloid precursor protein without decreasing Abeta production in human primary neuron cultures. J Neurosci, 1998. 18(8): p. 2907- 13. 69. Rossner, S., et al., Constitutive overactivation of protein kinase C in guinea pig brain increases alpha-secretoryAPP processing without decreasing beta-amyloid generation. Eur JNeurosci, 2000. 12(9): p. 3191-200. 70. Skovronsky, D.M., et al., Protein kinase C-dependent alpha-secretase competes with beta-secretase fo r cleavage of amyloid-beta precursor protein in the trans golgi network. J Biol Chem, 2000. 275(4): p. 2568-75. 71. Sinha, S., et al., Purification and cloning of amyloid precursor protein beta secretasefromhuman brain. Nature, 1999. 402(6761): p. 537-40. 72. Vassar,R., et al., Beta-secretase cleavage of Alzheimer's amyloid precursor protein by the transmembrane aspartic protease BACE. Science, 1999. 286(5440):
97 p. 735-41. 73. Yan,R., et al., Me mbrane-anchored aspartylprotease with Alzheimer's disease beta-secretase activity. Nature, 1999. 402(6761): p. 533-7. 74. Hong, L., et al., Structure of the protease domain of memapsin 2 (beta-secretase) complexed with inhibitor. Science, 2000. 290(5489): p. 150-3. 75. Capell, A., et al., Ma turation and pro-peptide cleavage of beta-secretase. J Biol Chem, 2000. 275( 40): p. 30849-54. 76. Cai, H., et al., BA CEJ is the major beta-secretase fo r generation of Abeta peptides by neurons. Nat Neurosci, 2001. 4(3): p. 233-4. 77. Luo, Y., et al., Mice deficientin BA CEJ, the Alzheimer's beta-secretase, have normal phenotype and abolished beta-amyloid generation. Nat Neurosci, 2001. 4(3): p. 231-2. 78. Sisodia, S.S. and P.H. St George-Hyslop, y-Secretase, Notch, Abeta and Alzheimer's disease: where do the presenilinsfitin? Nat Rev Neurosci, 2002. 3(4): p. 281-90. 79. Thinakaran,G., et al., Endoproteolysis of presenilin 1 and accumulation of processed de rivatives in vivo. Neuron, 1996. 17(1): p. 181-90. 80. De Strooper, B., et al., Deficiency of presenilin-1 inhibits the normal cleavage of amyloidprecursor protein. Nature, 1998. 391(6665): p. 387-90. 81. Herreman, A., et al., Tota l inactivation of y-secretase activity in presenilin deficientemb ryonic stem cells. Nat Cell Biol, 2000. 2(7): p. 461-2. 82. Zhang, Z., et al., Presenilins are requiredfo r y-secretase cleavage of beta-APP and transmembrane cleavage of No tch-I. Nat Cell Biol, 2000. 2(7): p. 463-5. 83. Wolfe,M.S., et al., Two transmembrane aspartates in presenilin-1 requiredfo r presenilin endoproteolysis and y-secretase activity. Nature, 1999. 398(6727): p. 513-7. 84. Kimberly, W.T., et al., The transmembraneaspa rta tes in presenilin 1 and 2 are obligatoryfo r y-secretase activityan d amyloid beta-protein generation. J Biol Chem, 2000. 275(5): p. 3173-8. 85. Leimer, U., et al., Zebrafish (Da nio rerio) presenilin promotes aberrant amyloid beta-peptide production and requires a critical aspartate residue fo r its fu nction in amyloidogenesis. Biochemistry, 1999. 38(41): p. 13602-9. 86. Steiner, H., et al., A loss of.function mutation of presenilin-2 interfe res with amyloid beta-peptide production and notch signaling. J Biol Chem, 1999. 274(40): p. 28669-73. 87. Esler, W.P ., et al., Tra nsition-state analogue inhibitors of y-secretase bind directly to presenilin-1. Nat Cell Biol, 2000. 2(7): p. 428-34. 88. Li, Y.M., et al., Photoactivated y-secretase inhibitors directed to the active site covalently label presenilin 1. Nature, 2000. 405(6787): p. 689-94. 89. Yu, G., et al., Nicastrin modulates presenilin-media ted notchlglp-1 signal transduction and betaAPPproc essing. Nature, 2000. 407(6800): p. 48-54. 90. Chung, H.M. and G. Struhl, Nicastrinis required/or Presenilin-mediated transmembranecl eavage in Drosophila. Nat Cell Biol, 2001. 3(12): p. 1129-32. 91. Hu, Y., Y. Ye, and M.E. Fortini, Nicastrin is required/or y-secretase cleavage of the Drosophila No tch receptor. Dev Cell, 2002. 2(1): p. 69-78.
98 92. Lopez-Schier, H. and D. St Johnston,Drosop hila nicastrinis essential fo r the intramembranous cleavage of notch. Dev Cell, 2002. 2(1): p. 79-89. 93. Li, T., et al., Nicastrinis required fo r assembly of presenilinly-secretase complexes to mediate Notch signalingan dfo r processing and traffickingof beta amyloid precursor protein in mammals. J Neurosci, 2003. 23(8): p. 3272-7. 94. Goutte, C., et al., APH- 1 is a multipass membrane protein essential fo r the Notch signaling pathway in Ca enorhabditis elegans embryos. Proc Natl Acad Sci U S A, 2002. 99(2): p. 775-9. 95. Francis, R., et al., aph-1 and pen-2 are requiredfo r Notch pathway signaling, y secretase cleavage of betaAPP, and presenilin protein accumulation. Dev Cell, 2002. 3(1): p. 85-97. 96. Lee, S.F ., et al., A conserved GXXXG motifin APH- 1 is critical fo r assembly and activityof the y-secretase complex. J Biol Chem, 2004. 279(6): p. 4144-52. 97. Takasugi, N., et al., The role of presenilin cofactors in the y-secretase complex. Nature, 2003. 422(6930): p. 438-41. 98. Edbauer, D., et al., Reconstitution of y-secretase activity. Nat Cell Biol, 2003. 5(5): p. 486-8. 99. De Strooper, B., et al., A presenilin-1-dependent y-secretase-like protease mediates release of Notch intracellulardo main. Nature, 1999. 398(6727): p. 518- 22. 100. Ni, C. Y., et al., y -Secretase cleavage and nuclear localization of ErbB-4 receptor tyrosinekinase. Science, 2001. 294(5549): p. 2179-81. 101. Scheinfeld, M.H., et al., Processing of beta-amyloid precursor-likepr otein-] and -2 by y-secretase regulates transcription. J Biol Chem, 2002. 277( 46): p. 44195- 201. 102. Marambaud,P ., et al., A CBP binding transcriptional repressor produced by the PSJ/epsilon-cleavage of N-cadherin is inhibited by PSJ FA D mutations. Cell, 2003. 114(5): p. 635-45. 103. Lammich, S., et al., Presenilin-dependent intramembrane proteolysis of CD44 leads to the liberation of its intracellular domain and the secretion of an Abeta like peptide. J Biol Chem, 2002. 277(47): p. 44754-9. 104. Brou, C., et al., A novel proteolytic cleavage involved in Notch signaling: the role of the disintegrin-metalloprotease TA CE. Mol Cell, 2000. 5(2): p. 207-16. 105. Hartmann, D., et al., The disintegrin/metalloprotease ADAM 10 is essential fo r Notch signalling but not fo r alpha-secretase activityin fibroblasts. Hum Mol Genet, 2002. 11(21): p. 2615-24. 106. Schroeter, E.H., J.A. Kisslinger, and R. Kopan, Notch-] signalling requires ligand-induced proteolytic release of intracellular domain. Nature,1998. 393(6683): p. 382-6. 107. Selkoe, D. andR. Kopan, Notch and Presenilin: regulated intramembrane proteolysis links development and degeneration. AnnuRev Neurosci, 2003. 26: p. 565-97. 108. Okochi, M., et al., Presenilins mediate a dual intramembranous y-secretase cleavage o/ Notch-1. Embo J, 2002. 21(20): p. 5408-16. 109. Sastre, M., et al., Presenilin-dependent y-secretase processing of beta-amyloid
99 precursor protein at a site corresponding to the S3 cleavage of Notch. EMBO Rep, 2001. 2(9): p. 835-41. 110. Weidemann, A., et al., A novel epsilon-cleavage within the transmembrane domain of the Alzheimer amyloid precursor protein demonstrates homology with Notch processing. Biochemistry, 2002. 41(8): p. 2825-35. 111. Yu, C., et al., Characterization of a presenilin-mediated amyloid precursor protein carboxyl-terminal fragment y. Evidence fo r distinct mechanisms involved in y -secretase processing of the APP and Notch] transmembrane domains. J Biol Chem, 2001. 276(47): p. 43756-60. 112. Cao, X. andT.C. Sudhof, A transcriptionally [correction of transcriptively] active complex of APP with Fe 65 and histone acetyltransferase Tip 60. Science, 2001. 293(5527): p. 115-20. 113. Petit, A., et al., New protease inhibitors prevent y-secretase-mediated production of Abeta40/4 2 without affecting Notch cleavage. Nat Cell Biol, 2001. 3(5): p. 507- 11. 114. Esler, W.P., et al., Amyloid-lowering isocoumarins are not direct inhibitors of y secretase. Nat Cell Biol, 2002. 4(5): p. El 10-1; author reply El 11-2. 115. Weggen, S., et al., A subset of NSAIDs lower amyloidogenic Abeta42 independently of cyclooxygenase activity. Nature, 2001. 414(6860): p. 212-6. 116. Zhou, Y., et al., Nonsteroidal anti-inflammatorydr ugs can lower amyloidogenic Abeta42 by inhibiting Rho. Science, 2003. 302(5648): p. 1215-7. 117. Hersh, L.B. and K. Morihara, Comparison of the subsite specificityof the mammalian neutral endopeptidase 24. 11 (e nkephalinase) to the bacterial neutral endopeptidase thermolysin. J Biol Chem, 1986. 261(14): p. 6433-7. 118. Howell, S., J. Nalbantoglu, and P. Crine, Neutral endopeptidase can hydrolyze beta-amyloid(J-40) but shows no effect on beta-amyloid precursor protein metabolism. Peptides, 1995. 16(4): p. 647-52. 119. Shirotani, K., et al., Neprilysin degrades both amyloid beta peptides 1-40 and 1- 42 most rapidly and effi ciently among thiorphan- and phosphoramidon-sensitive endopeptidases. J Biol Chem, 2001. 276(24): p. 21895-901. 120. Iwata, N., et al., Me tabolic regulation of brain Abeta by neprilysin. Science, 2001. 292(5521): p. 1550-2. 121. Rama, E., et al., Clearance of extracellular and cell-associated amyloid beta peptide through viral expression of neprilysin in primaryneurons. J Biochem (Tokyo), 2001. 130(6): p. 721-6. 122. Y asojima, K., et al., Reduced neprilysin in high plaque areas of Alzheimer brain: a possible relationship to deficientdegrad ation of beta-amyloid peptide. Neurosci Lett, 2001. 297(2): p. 97-100. 123. Kurochkin, I.V. and S. Goto, Alzheimer's beta-amyloid peptide sp ecifically interacts with and is degradedby insulin degrading enzyme. FEBS Lett, 1994. 345(1): p. 33-7. 124. Qiu, W.Q., et al., Insulin-degrading enzyme regulates extracellular levels of amyloid beta-protein by degradation. J Biol Chem, 1998. 273(49): p. 32730-8. 125. Farris, W., et al., Insulin-degrading enzyme regulates the levels of insulin, amyloid beta-protein, and the beta-amyloid precursor protein intracellular
100 domain in vivo. Proc Natl Acad Sci US A, 2003. 100(7): p. 4162-7. 126. Ertekin-Taner, N., et al., Linkage of plasma Abeta42 to a quantitative locus on chromosome 10 in late-onset Alzheimer's disease pedigrees. Science, 2000. 290(5500): p. 2303-4. 127. Hu, J., et al., Angiotensin-converting enzymegenoty pe is associated with Alzheimer disease in the Japanese population. Neurosci Lett, 1999. 277(1): p. 65- 7. 128. Eckman, E.A., D.K. Reed, and C.B. Eckman, Degradation of the Alzheimer's amyloid beta peptide by endothelin-converting enzyme. J Biol Chem, 2001. 276(27): p. 24540-8. 129. Kingston, LB., M.J. Castro, and S. Anderson, In vitrostimulation of tissue-type plasminogen activator by Alzheimer amyloid beta-peptide analogues. Nat Med, 1995. 1(2): p. 138-42. 130. Kimberly, W.T. and M.S. Wolfe,Ide ntity andfu nction of y-secretase. J Neurosci Res, 2003. 74(3): p. 353-60. 131. Gu, Y., et al., Distinct intramembrane cleavage of the beta-amyloid precursor protein fa mily resembling y-secretase-like cleavage of Notch. J Biol Chem, 2001. 276(38): p. 35235-8. 132. Kim, S.H., et al.,Mu ltiple effects of aspartate mutant presenilin 1 on the processing and trafficking of amyloid precursor protein. J Biol Chem, 2001. 276(46) : p. 43343-50. 133. Chen, F., et al.,Presenilin 1 mutations activate y 42-secretase but reciprocally inhibit epsilon-secretase cleavage of amyloid precursor protein (A PP) and S3- cleavage of notch. J Biol Chem, 2002. 277(39): p. 36521 -6. 134. Ikeuchi, T., et al., Familial Alzheimer disease-linked presenilin 1 variants enhance production of both Abeta 1-40 and Abeta 1-42 peptides that are only partially sensitive to a potent aspartylprotease transition state inhibitor of "y secretase ". J Biol Chem, 2003. 278(9): p. 701 0-8. 135. Wiltfang, J., et al., Improved electrophoretic separation and immunoblotting of beta-amyloid (A beta) peptides 1-40, 1-42, and 1-43. Electrophoresis, 1997. 18(3- 4): p. 527-32. 136. Kuo, Y.M., et al., Elevated abeta42 in skeletal muscle of Alzheimer disease patients suggests peripheralalterations of AbetaPP metabolism. Am J Pathol, 2000. 156(3): p. 797-805. 137. Roher, A.E., et al., Thehuman amyloid-beta precursor protein770 mutation V71 7F generates peptides longer than amyloid-beta-(40-42) andflocculent amyloid aggregates. J Biol Chem, 2004. 279(7): p. 5829-36. 138. Hardy, J., Framing beta-amyloid Nat Genet, 1992. 1(4) : p. 233-4. 139. Zhao, G., et al., Identification of a new presenilin-dependent zeta-cleavage site within the transmembrane domain of amyloid precursor protein. J Biol Chem, 2004. 279(49): p. 50647-50. 140. Thinakaran, G., et al., Me tabolism of the "Swedish " amyloid precursor protein variant in neuro2a (N2a) cells. Evidence that cleavage at the "beta-secretase " site occurs in the golgi apparatus. J Biol Chem, 1996. 271(16): p. 9390-7. 141. Pinnix, I., et al., A novel y -secretase assay based on de tection of the putative C-
101 terminalfragment-y of amyloid beta protein precursor. J Biol Chem, 2001. 276(1): p. 481-7. 142. Farmery,M.R., et al., Partial purificationand characterization of y-secretase frompost-mortem human brain. J Biol Chem, 2003. 278(27): p. 24277-84. 143. Li, Y.M., et al., Presenilin 1 is linked with y-secretase activity in the detergent solubilized state. Proc Natl Acad Sci USA, 2000. 97(11): p. 6138-43. 144. Xu, X., et al., The novel presenilin-1-associated protein is a proapoptotic mitochondrial protein. J Biol Chem, 2002. 277(50): p. 48913-22. 145. Xia, W., et al., Presenilin complexes with the C-terminal fragments of amyloid precursor protein at the sites of amyloid beta -protein generation [In Process Citation]. Proc Natl Acad Sci U S A, 2000. 97(16): p. 9299-304. 146. Domin, J., et al., Th e class IIphos phoinositide 3-kinase P/3K-C2alpha is concentrated in the trans-Golgi network and present in clathrin-coated vesicles. J Biol Chem, 2000. 275(16): p. 11943-50. 147. Funamoto,S., et al., Tr uncated Carboxyl-Terminal Fragments of beta-Amyloid Precursor Protein Are Processed to Amyloid beta-Proteins 40 and 42. Biochemistry, 2004. 43(42): p. 13532-40. 148. Tian, G., et al., The mechanism of y-secretase: multiple inhibitor binding sites fo r transitionstate analogs and small molecule inhibitors. J Biol Chem, 2003. 278(31 ): p. 28968-75. 149. Esler, W.P., et al., Activity-dependent isolation of the presenilin- y -secretase complex reveals nicastrin and a y substrate. Proc Natl Acad Sci U S A, 2002. 99(5): p. 2720-5. 150. Seiffert, D., et al., Presenilin-1 and -2 are molecular targets fo r y-secretase inhibitors. J Biol Chem, 2000. 275(44): p. 34086-91. 151. Citron,M., [beta]-Secretase inhibition fo r the treatmentof Alzheimer's disease - promise and challenge. 2004. 25(2): p. 92. 152. Pollack, SJ. and H. Lewis, Se cretase inhibitors fo r Alzheimer's disease: challenges of a promiscuous protease. Curr Opin Investig Drugs, 2005. 6(1): p. 35-47. 153. Higaki, J., et al., Inhibition of beta-amyloidformation identifiesproteol ytic precursors and subcellular site of catabolism. Neuron, 1995. 14(3): p. 651-9. 154. Wolfe, M.S., et al., A substrate-based difluoro ketone selectively inhibits Alzheimer's y-secretase activity. J Med Chem, 1998. 41(1): p. 6-9. 155. Wolfe, M.S., et al., Peptidomimetic probes and molecular modeling suggest that Alzheimer'sy-secret ase is an intramembrane-cleaving aspartylprotease. Biochemistry, 1999. 38(15): p. 4720-7. 156. Dovey, H.F., et al., Functional y-secretase inhibitors reduce beta-amyloid peptide levels in brain. JNeurochem, 2001. 76(1): p. 173-81. 157. Shearman, M.S., et al., L-685, 458, an aspartylprotease transition state mimic, is a potent inhibitor of amyloid beta-protein precursor y-secretase activity. Biochemistry, 2000. 39(30): p. 8698-704. 158. Lanz, T.A., et al., The y-secretase inhibitor N- [N- (3,5-difluorophenacetyl)-L alanyl]-S-phenylglycine t-butyl esterreduces A beta levels in vivo in plasma and cerebrospinalfluidin young (plaque-free) and aged (plaque-bearing) Tg2576
102 mice. J Pharmacol Exp Tuer, 2003. 305(3): p. 864-71. 159. Tsai, J.Y., M.S. Wolfe,and W. Xia, The search/or y-secretase and development of inhibitors. CurrMed Chem, 2002. 9(11): p. 1087-106. 160. Beher, D., et al., Pharmacological knock-down of the presenilin 1 heterodimer by a novel y -secretase inhibitor: implications fo r presenilin biology. J Biol Chem, 2001. 276(48): p. 45394-402. 161. Gouras, G.K., et al., Intraneuronal Abeta42 accumulation in human brain. Am J Pathol, 2000. 156(1): p. 15-20. 162. LaFerla, F.M., et al., The Alzheimer's A beta peptide induces neurodegeneration and ap optotic cell death in transgenic mice. Nat Genet, 1995. 9(1): p. 21-30. 163. Takahashi, R.H., et al., Intraneuronal Alzheimer abeta42 accumulates in multivesicular bodies and is associated with synaptic pathology. Am J Pathol, 2002. 161(5): p. 1869-79. 164. Magrane, J., et al., He at shock protein 70 participates in the neuroprotective response to intracellularly expressed beta-amyloid in neurons. J Neurosci, 2004. 24(7): p. 1700-6. 165. Zhang, Y., et al., Selective cytotoxicityof intracellular amyloid beta peptidel-42 through p53 and Bax in cultured primaryhuman neurons. J Cell Biol, 2002. 156(3): p. 519-29. 166. Rishton, G.M., et al., Fenchylamine sulfonamide inhibitors of amyloid beta peptide production by the y-secretase proteolytic pathway:potential small molecule therapeutic agents fo r the treatmentof Alzheimer's disease. J Med Chem, 2000. 43(12): p. 2297-9. 167. Hartmann, T., et al., Distinct sites of intracellularprod uction fo r Alzheimer's disease A beta40/42 amyloidpeptides. Nat Med, 1997. 3(9): p. 1016-20. 168. Sinha, S. andI. Lieberburg,Cellular mechanisms of beta-amyloid production and secretion. Proc Natl Acad Sci USA, 1999. 96(20): p. 11049-53. 169. Selkoe, D.J. andM.B. Podlisny, Deciphering the genetic basis of Alzheimer's disease. Annu Rev Genomics Hum Genet, 2002. 3: p. 67-99. 170. Mehdi, S., et al., Inhibition of the proteolysis of rat erythrocyte membrane proteins by a synthetic inhibitor of ca/pain. Biochem Biophys Res Commun, 1988. 157(3): p. 1117-23. 171. Harrison, T., I. Churcher, and D. Beher, y-Secretase as a targetfo r drug intervention in Alzheimer's disease. Curr Opin Drug Discov Devel, 2004. 7(5): p. 709-19. 172. Annaert, W.G., et al., Presenilin 1 controls y-secretase processing of amyloid precursor protein in pre-golgi compartments of hippocampal neurons. J Cell Biol, 1999. 147(2): p. 277-94. 173. Culvenor, J.G., et al., Alzheimer's disease-associated presenilin 1 in neuronal cells: evidence fo r localization to the endoplasmic reticulum-Golgi intermediate compartment. J Neurosci Res, 1997. 49(6): p. 719-31. 174. Kim, S.H., et al., Subcellular localization of presenilins: association with a unique membrane pool in cultured cells. Neurobiol Dis, 2000. 7(2): p. 99-117. 175. Walter, J., et al., Th e Alzheimer's disease-associated presenilins are differentially phosphorylatedproteins located predominantly within the endop lasmic reticulum.
103 Mol Med, 1996. 2(6): p. 673-91. 176. De Strooper, B., et al., Study of the synthesis and secretion of normal and artificialmu tants of murine amyloidprecursor protein (A PP): cleavage of APP occurs in a late compartment of the default secretion pathway. J Cell Biol, 1993. 121(2): p. 295-304. 177. Koo, E.H. and S.L. Squazzo, Evidence that production and release of amyloid beta-protein involves the endocytic pathway. J Biol Chem, 1994. 269(26): p. 17386-9. 178. Selkoe, D.J., The cell biology of beta-amyloid precursor protein and presenilin in Alzheimer's disease. Trends Cell Biol, 1998. 8(1 1): p. 447-53. 179. Sisodia, S.S., Beta-amyloid precursor protein cleavage by a membrane-bound protease. Proc Natl Acad Sci USA, 1992. 89(13): p. 6075-9. 180. Cook, D.G., et al., Alzheimer's A beta(l-42) is generated in the endoplasmic reticulum/intermediate compartment of NT2N cells. Nat Med, 1997. 3(9): p. 1021- 3. 181. Wild-Bode, C., et al., Intracellular generation and accumulation of amyloid beta peptide terminating at amino acid 42. J Biol Chem, 1997. 272(26): p. 16085-8. 182. Greenfield,J.P., et al., Endoplasmic reticulum and trans-Golgi network generate distinct populations of Alzheimer beta-amyloid peptides. Proc Natl Acad Sci U S A, 1999. 96(2): p. 742-7. 183. Xu, H., et al., Generation of Alzheimer beta-amyloid protein in the trans-Golgi network in the apparent absence of vesicle fo rmation. Proc Natl Acad Sci U S A, 1997. 94(8): p. 3748-52. 184. Chyung, J.H. and D.J. Selkoe, Inhibition of receptor-mediated endocytosis demonstratesgeneration of amyloid beta-protein at the cell surface. J Biol Chem, 2003. 278(5 1): p. 51035-43. 185. Haass, C., et al., Targeting of cell-surfa ce beta-amyloid precursor protein to lysosomes: alternativeprocessing into amyloid-bearingfragments. Nature, 1992. 357(6378): p. 500-3. 186. Grimm, H.S., et al., y-Secretase cleavage site sp ecificitydiff ers fo r intracellular and secretoryamylo id beta. J Biol Chem, 2003. 278(15): p. 13077-85. 187. Martin, B.L., et al., Intracellular accumulation of beta-amyloid in cells expressing the Swedish mutant amyloid precursor protein. J Biol Chem, 1995. 270(45): p. 26727-30. 188. Caporaso, G.L., et al., Chloroquine inhibits intracellular degradation but not secretion of Alzheimer beta/A4 amyloid precursor protein. Proc Natl Acad Sci U S A, 1992. 89(6): p. 2252-6. 189. Gabuzda, D., et al., Inhibition of energymet abolism alters the processing of amyloid precursor protein and induces a potentially amyloidogenic derivative. J Biol Chem, 1994. 269(18): p. 13623-8. 190. Mollenhauer, H.H., D.J. Morre, and L.D. Rowe, Alteration of intracellular traffic by monensin; mechanism, sp ecificityand relationship to toxicity. Biochim Biophys Acta, 1990. 1031(2) : p. 225-46. 191. Tartakoff, A.M., Perturbation of vesicular traffic with the carboxylic ionophore monensin. Cell, 1983. 32(4): p. 1026-8.
104 192. Bole, D.G., et al., Im munocytochemical localization of BiP to the rough endoplasmic reticulum: evidence fo r protein sorting by selective retention. J Histochem Cytochem, 1989. 37(12): p. 1817-23. 193. Melnick, J., J.L. Dul, and Y. Argon, Sequential interaction of the chaperones BiP and GRP94 with immunoglobulin chains in the endoplasmic reticulum. Nature, 1994. 370(6488): p. 373-5. 194. Robinson, M.S., Cloning and expression of y-adap tin, a component of clathrin coated vesicles associated with the Golgi apparatus. J Cell Biol, 1990. 111 (6 Pt 1 ): p. 2319-26. 195. Caporaso, G.L., et al., Mo rphologic and biochemical analysis of the intracellular traffickingof the Alzheimer beta/A 4 amyloid precursor protein. J N eurosci, 1994. 14(5 Pt 2): p. 3122-38. 196. Kuentzel, S.L., et al., The Alzheimer beta-amyloid protein precursor/p rotease nexin-IIis cleaved by secretase in a trans-Golgi secretorycom partment in human neuroglioma cells. Biochem J, 1993. 295 (Pt 2): p. 367-78. 197. Sambamurti, K., et al., Evidence fo r intracellular cleavage of the Alzheimer's amyloid precursor in PCJ 2 cells. J Neurosci Res, 1992. 33(2): p. 319-29. 198. Fenteany, G., et al., Inhibition ofproteasome activities and subunit-specific amino-terminal threonine modification by lactacystin. Science, 1995. 268(5211): p. 726-31. 199. Drexler, H.C., The role ofp 27Kip l inproteasome inhibitor induced ap optosis. Cell Cycle, 2003. 2(5): p. 438-41. 200. Golde, T.E., et al., Processing of the amyloidprotein precursor to potentially amyloidogenic derivatives. Science, 1992. 255(5045): p. 728-30. 201. Pahlsson,P. and S.L. Spitalnik, The role of glycosylation in synthesis and secretion of beta-amyloid precursor protein by Chinese hamster ovarycells. Arch Biochem Biophys, 1996. 331(2): p. 177-86. 202. Chyung, A.S., et al., Novel beta-secretase cleavage of beta-amyloid precursor protein in the endoplasmic reticulum/intermediate compartmentof NT2N cells. J Cell Biol, 1997. 138(3): p. 671-80. 203. Iwata, H., et al., Subcellular compartment and molecular subdomain of beta amyloidprecursor protein relevant to the Abeta 42-promoting effects of Alzheimer mutant presenilin 2. J Biol Chem, 2001. 276(24): p. 21678-85. 204. Maltese, W.A., et al.,Retention of the Alzheimer's amyloid precursor fragment C99 in the endoplasmic reticulum prevents fo rmation of amyloid beta-peptide. J Biol Chem, 2001. 276(23): p. 20267-79. 205. Sudoh, S., et al., Intracellular site of y-secretase cleavage fo r Abeta42 generation in neuro 2a cells harbouring a presenilin I mutation. Eur J Biochem, 2000. 267(7): p. 2036-45. 206. Cosson, P. andF. Letourneur, Coatomer interaction with di-lysine endoplasmic reticulum retention motifs. Science, 1994. 263(5153): p. 1629-31. 207. Orci, L., et al., Buddingfrom Golgi membranes requires the coatomer complex of non-clathrin coat proteins. Nature, 1993. 362(6421): p. 648-52. 208. Peraus, G.C., C.L. Masters, and K. Beyreuther, Late compartments of amyloid precursor protein transport in SY5 Y cells are involved in beta-amyloid secretion. 105 JNeurosci, 1997. 17(20): p. 7714-24. 209. Ponnambalam, S., et al., The TGN38 glycoprotein contains two non-overlapp ing signals that mediate localization to the trans-Golgi network. J Cell Biol, 1994. 125(2): p. 253-68. 210. Cupers, P., et al., Th e discrepancybetween presenilin subcellular localization and y-secretase processing of amyloid precursor protein. J Cell Biol, 2001. 154(4): p. 731-40. 211. Kaether, C., et al., Presenilin-1 affects traffickingand processing of betaAPP and is targeted in a complex with nicastrin to the plasma membrane. J Cell Biol, 2002. 158(3): p. 551-61. 212. Ray, W.J., et al., Cellsur fa ce presenilin-1 participates in the y-secretase-like proteolysis of Notch. J Biol Chem, 1999. 274(51): p. 36801-7.
106 APPENDIX
107 Ta bl e 1 ·,-secret ase m. h"b"t1 1 or m. tiorma ion f Name Synonyms Structure or sequence Fianl concentration y40-I -3,5-DMC-IL-CHO trans-3,5-DMC-Ile-Leu-CHO 20 µM y40-II N-ter-Butyloxycarbonyl-Gly-Val- BOC-Gly-Val-Val-CHO lO0 µM Valinal y-lV 2-Naphthoyl-VF-CHO 2-Naphthoyl-Val-Phe-CHO 25 µM y-V Z-LF-CHO Z-Leu-Phe-CHO 25 uM y-VI 1-(S}-endo-N-(1,3,3)- q(i � 20 µM Trimethylbiocyclo[2.2.1] hept-2-yl}-4-fluorophenyl -� � H C r:J Sulfonamide 3 8--QF
y-IX DAPT 0.5 µM
·�0 �0�I-¼ !20r••" F
y-X 1685,458 0.5 µM --��Cr O,u� ,i:;2-. - ' � ""'-·t:_:::i X 0 � y-XI JLK6 N 100 µM O � CH3 Cl O y-XII Z-IL-CHO Z-Ile-Leu-CHO 40 uM y-XIII Z-YIL-CHO Z-YIL-CHO 50 µM y-XIV Z-C(t-Bu}-IL-CHO Z-Cys( t-Bu }-Ile-Leu-CHO 1.0 µM
y-XVI DAPM 0.1 µM
' �0: y1! y-XVII 31C, WPE-III-3 1C l0 µM r N ° � >(ogl;l l'-NA�__,Jl� . o--.. D l_r� 0 ()I<. y-XVIII Compound E t 3nM -� -2-�t y-XIX (2S,3R}-3-(3,4-Difluorophenyl)- 3nM 2-(4-fluorophenyl}-4-hydroxyl-N- ((3S)-2-oxo-5-phenyl-2,3- dihydro-lH-benzo[e][l,4] �� ��: diazepin-3-yl)-butyramide Ibuprofen 2-( 4-Isobutylphenyl)-propionic Ha\-� ,CH, 500 µM Acid C \ CH H3C J C02H Sulind Z-5-Fluoro-2-methyl- H3C- 60 µM sulfide 1-[p-(methylthio) benzylidene] � indene-3-acetic acid CH3 F I COOH 108 TM APP A� N + 'V40++ 'V4? ISEVKMDAEFRHDSGYEVHHQKLVFF AEDVGSNKGAIIGLMVGGVVIA TVIVITL VMLKK ... NL GQN IMVF P K G G Swedish L -...,l APP mutations (Red) -...,l London& Indiana Chapter 1 Fig.1. Some known mutations in APP [1 ]. 109 - ,•. •. •• �t�� 11····.. · .··•ntr..·v 2 3 4 5 CTFa(end) �Q �TFJ3 � ., \0 A�40 42 � ,..... �f3 � -A�43 ,... lJII ·.·· 1n1 . 3 4 5 6 AA46J 2 3 4 5 6 7 t- A B Chapter 2 Fig.1. Detection of a new intracellular AfJ species. APP derivatives in lysates or media of cells treated with (lanes 2 to 5) or without (lane 1) various inhibitors (500 nM DAPT, 100 nM DAPM, 3 nM compound E (CPDE), and 5 nM XIX, respectively) for 8hrs were analyzed by 10- 18% Glycineffris-SDS-PAGE (top panels in both Fig I A and B) or by Bicine/urea SDS-PAGE (bottom panels in both Fig 1 A and B). Lane 6 is the mix of synthetic A�40 and A�42. Lane 7 is the synthetic A�43. The top panel in Figl A is the blot probed with APP C-terminal specificantibody Cl 5, the other three panels were probed with 6El0. Since the CTF� and CTFa derived from recombinant APPsw are tagged with a myc epitope, they were detected as bands with slower migration rates than the endogenous CTFa (CTFa(end))(top pane). The bottom panel in Fig 1 A is the secreted A� immunoprecipitated from CM. 110 HEK293 HS683 Mt 7 N2aAPPmyc + + - + - + CPDE --fAPP �CTEB_--',myc ---CTFJ3 --A 40 --A;,42 �A 46_ l 2 3 4 S 6 7 8 9 Chapter 2 Fig.2. Detection of the novel AfJspecies from several cell lines. Cell lysates of HEK293 (lanesl,2), human glioma HS683 (lanes 3 and 4), and humanneuroblastoma Ml 7 (lanes 5 and 6) cells treated with (3 nM) or without (-) compound E (CPDE) as indicated for 8 hrs were analyzed for the presence of the new A� species. As controls, cell lysates of N2a cells stably expressing APPsw and PS 1 wt treated with (lane 8) or without (lane 7) compound E were included. Lane 9 is the mix of synthetic A�40 and A�42. 111 Voya,g41rlpt¢-t1= >NF0,7=>MC--,,m:::�ec=,.RIMSOO=,.SC-:>TR=>-BC[BP = ,, 21.2, 12:7J 4!!27.53 t I Chapter 2 Fig.3. MALDI-MS spectrum forthe novel AP species. Mass spectrometric analysis was performed as described previously[133] except that after applying the sample matrix solution to the sample plate, formic acid and isopropanol were added to a finalconcentration of 30% each. 112 fPS1 fA PP- AJ346- sAPP- Af3- Chapter 2 Fig.4. Dominant negative form of PSI diminishes Ap46 generation. Cell lines are labeled at the top. Lysates of cells cultured in the presence (+) or absence (-) of 0.5 µM DAPT were analyzed by Western blot forthe expression of PSl(top panel) using anti-PSlN, which is specific for the N-terminal domain of PSl and was raised against a peptide of 24 amino acids corresponding to residues 27-50 of human PSl. Full-length APP, CTF�, and A�46 in the cell lysate were detected by 6E10 (middle panel). Secreted A� and soluble APP (sAPP) immunoprecipitated from CM were also detected by 6E10 (bottompane l). The two bands between sAPP and A� are the heavy chain and light chain of IgG (bottompane l). 113 PS1PS2knockout PSl/PSlWt N2a PS1 PS2knockout PSl/PS2Wt APPsw + + • - + + + APPsw - - + + . - + + Lael CPD E Lael + + . + + i CPD E - + - + + - + Ul. < 1\4(),'42 A 3 4 5 6 7 B 9 10 B 2 3 4 5 6 7 8 9 10 Chapter 2 Fig.5. Double knockout of PS1/PS2 abolishes AfJ46 generation. PS 1/PS2 wild-type or double knockout cells were transiently transfected with either Lacz or APPsw vectors and cultured in the presence or absence of compound E for 12 h. The lysates were separated either on 10/18% SDS-PAGE (top panel) or on 10%Bicine/urea SDS-PAGE and probed with 6E10 antibody. Note that A�46 was only seen in wild-type (lane8) but not PS double knockout cells (lane4). 114 r L685(µM) -, r31C(µM>, rDAPT(µM) 7 r DAPM(nM)7 i rCPDE(nM) 7r XIX(nM) 7 ; Ctr0.02 0.1 0.5 2.5 4 20 100 500 � Ctr 0.2 1 5 25 0.2 1 5 25 ... � -- - --�--·. . ,. . n•:,t ·fAPP , -- ....., _:...., . �-..;.� ·� #'.<··· � 11a till . -- _..,. -CTFP =��g ----- · - i• illl ,· ·AJl46 i·r:·�2Ll: :·1:1rn��tl·-·�·-'C:�� Chapter 2 Fig.6. Deferential inhibitions of AP46 formation by transition state analogs and nontransition state inhibitors. Accumulation of CTF� and the new intracellular A�46 (upper panel) and the secreted A�40 and A�42 immunoprecipitated from CM with 6E 10 (lower panel) were analyzed by 10% Bicine/urea PAGE followedby Western blotting using 6E 10. Inhibitors and concentrations used are indicated on the top of the gels. Lanes 1, 9, and 19 are controls of samples fromuntreated cells. Lanes 18 and 28 arethe mix of synthetic A�0 and A�42 used as standards. 115 CPD E I fAPP CTF� ---AJ340 1 A�42 ·A,�46 -�A�40 -AP42 Chapter 2 Fig.7. The Indiana mutation of APP enhances AIJ46 generation. N2a cells were transiently transfected either with APPsw695 or with APPswN717F vectors and cultured in the presence or absence of 3 nM compound E. Lysates (top panel) and immunoprecipitated secreted A� species from CM (bottom panel) were separated on 10% Bicine/urea SDS-PAGE and probed with 6E10. Lane 5 is the mixture of synthetic A�40/42. 116 - 0.5 1.5 2.5 1.5 2.5 1.5 2.5 L685(µM) 5- -- -55 CPDE(nM) 0.5 - 0.5 0.5 DAPT(µM) Ctr --- 4 20 1 oo 500 DAPM(nM) -fAPP -CTFJ3 1�-A640 ,f-AJ346-Af342 Chapter 3 Fig.1. The absence of AP46 in cells treated with L- 685,458 is due to its inhibition of the formation of AJ346. Both cell lysates (top panels) and the secreted A�40/42immunoprecipitated from conditioned medium (CM) (bottom panels) were analyzed by 10% Bicine/urea-SDS-PAGE. Lane 1 is the control (Ctr) of cells cultured in absence of any inhibitors. Lane 15 is the mix of synthetic A�O and A�42. 117 �A p46 l 2 3 4 5 6 '7 · S 9 10 H Incubation of Membranefor 1 h Chapter 3 Fig.2. AP46 turnover in a cell-free system. Membranes were prepared as described in the "Experimental Procedures". Membranes were incubated at 37 °C or 0°C for 1 hr. Samples were analyzed by 10-18% regular SOS-PAGE followed by Western blotting using 6E 10. Lanes 1-4 were one batch and lanes 5-11 were another batch. 118 11 hr L DM - L 11- L DM - L 2 hrl A 40 min L DM DM DM/L - L DM DM DM/L � - 12 hr DM DM DM - - DM p_rv, D_rv1, ,· ..� 4 5 6 L-11 2 hou 3 r----J L-27 hou 8 rs 9__J 10 11 Chapter 3 Fig.3. Generation of secreted AfU0/42 from Ap46. L, L-685,458; DM, DAPM; upper panel, cell lysates; bottom panel, A� immunoprecipitated from CM; lane 11, the mix of synthetic A�40 and A�42. 119 Cell Lysates g n P(800 xg) S(PNS) .t;g S(20k X g) +P(20 k X g) Solublization t2ok x g t + SS(20k x g) SP(20k+ x g) lmmunoprecipitation Discard Chapter 3 Fig. 4. Schematic diagram of the procedure of crude membrane preparation and solubilization 120 a 0) )( � � C) -C, � � )( )( 1Q. "' a.. KDa 8 fAPP- � KDa \-100 ..36 -75 . 27 CTF�, 17 �8.8 A 6- 3.9 1 2 3 Fig. SA Fig. SB Cl> E � 1ii .s w � .� q. KDa 0-, fAPP _.:.J _ __ 100 75 -27 fPS1 '50 CTF13 -·--- 17 37 -8.8 AP4s..-•.__ 3. . 9 1 2 3 4 5 6 PS1 L- l GOPDE iH [�85 ! Fig. SC Fig. SD Chapter 3 Fig.S. AP46 is associated with PSl. Cell lysate fractionation and co immunoprecipitation were performed as illustrated in Fig.4 and described in "Materials and methods". The resulting fractions and immunoprecipitates were analyzed by 10-18% SDS-PAGE forA�46 (Fig. 5A and. 5C) and forthe C-terminal fragment of PSl (PSlL) (Fig. 5D) or 10% SDS-PAGE for PSlN (Fig. 5B) followed by western blotting. A�46 and CTF� were detected with 6El0. In Fig. 5B, PSl was detected with anti-PSlN, which recognizes the N-terminus of PSl [139]. Lane 1 in Fig. 5C is the synthetic A�46. In Fig. 5D, PSl in the immunoprecipitates was detected with anti-PSlL, which recognizes theC terminus of PSl [144]. Since both anti-PSlN and anti-PSlL are rabbit polyclonal antibodies, the rabbit IgG was detected as heavy dark bands above the PS 1 L band in lanes 2 and 3 of Fig. 5D. 121 � i. N � � �- (0 - � � r Lael , rAPPsw645, � rA PPsw6457 8:: -8: � Ctr E L Ctr E L �Ctr E l < <( CTF� .·.·=�&:: 1 2 3 4 5 6 7 8 9 10 A B Chapter 3 Fig.6. Detection of AP49 in intact cells and the production of secreted AP40/42from AP49. A. Detection of A�49 in intact cells. N2a cells, which stably express PS 1 wt, were transiently transfected with Lacz (lanes 1 to 3) or with APPsw645 (lanes 4 to 6), and the formation and turnover of A�49 and A�46 in the absence of inhibitors (Ctr) or in the presence of compound E (E) or L-685,458 (L) were analyzedby 10-18% regular SDS-PAGE followed by Western blotting using 6E10. Lanes 7 and 8 are Immunoprecipitates (IP) from lysates of N2a cells stably expressing APPsw and PS1 wt. Note, IP* indicates that immunoprecipitation was carried out in the presence of DAPT. Lane 9 is the synthetic A�46. Lane 10 is the synthetic A�49. The synthetic A�49 was used as a crude product because the purificationwas not successful, maybe due to its high hydrophobicity. Fig.6B. Production of secreted A�40/42from A�49. N2a cells stably expressing PS 1 wt were further transfected with APPsw645. Stable clones were selected with both 0418 (for APPsw) and Hygromycin (for presenilin). Secreted A�40/42in CM of both transiently transfected (lanes 2) and stably transfected cells (lane 5 and 6) was analyzed by 10% Bicine/urea SDS-PAGE. Lane 1 and lanes 7 are thesynthetic A�40/42. Since 6E 10 was used for both immunoprecipitation and Western blotting, the dark bands in Fig. 6A (lanes 7and 8, between CTF� and fAPP bands) and in Fig. 6B (lane 2 to 6, between A�40/42and sAPP bands) are the heavy chain and light chain of mouse IgG. 122 NH2 COOH +.11¾.l��-\l.�f�l�ltt��f:-�1«4 pi DAPT ------J Ef ..._--- L-685,458 DAPT ------1 l;t ....,_----L-685,458 ----�• DAPT ---.....I • Yt I Chapter 3 Fig.7. Schematic illustration of APP processing. L-685,458 specifically inhibits e- and �-cleavages (-); DAPT strongly and preferentially inhibits y-cleavage (-) and also inhibits e- and �-cleavages at a higher concentration (---). 123 -CTfp. t 3 4 6 7 8 --CTFu B 1 Chapter 4 Fig.I. Effects of DAPT, compound E, ibuprofen, and sulindac sulfide on AP46 and CTFP accumulation. N2aAPPsw/PS 1 wt cells were cultured in the presence or absence of y-secretase inhibitors as indicated for 10 h. Lysates were separated on 1 O/l 8%SDS-P AGE system. Panel A, top panel was immunoblotted with 6E 1 O; bottom panel was the longer exposure of top panel. Panel B. immuno-blotted with C 15. Ctrs represents control; CpdE, lbu, SSide stand for Conpound E, Ibuprofen, and Sulindac sulfide,repre spectively. 124 - = i I att2: >>� >< ><�� ����! 2 10 It 2 f 9 10 tJ ll 13 A , ; B --1 ----2 3 ---- 4 S -· 6 ·· 7 · 8,., · ...... 9 10 1'1 .....· 12 .. 13 ,14..... IS 16 = > = ;::, :s l - = i::ll e t � > > � >< < e x � x >< � S< $< e V, -A 40 -A�2 3 4 6 7 1:5 l 17 t8 l9 C Chapter 4 Fig.2. Inhibition profiles of various y-secretase inhibitors on APP processing. N2aAPPsw/PS1 wt cells were cultured in the presence various inhibitors as indicated in the figure for 10 hat concentration listed in Tabl except for 31C. A. Effects of various y-secretase inhibitors on intracellular accumulation of A�46 and CTF� . Lysates were analyzed by 10/18% SDS-PAGE and immunoblotted with 6E10( top panel), and C-l 5(bottom panel), respectively. B. Effects of various y-secretase inhibitors on soluble APP production. Conditioned media were separated on 8% SDS-PAGE, and blotted with 22Cll (top panel), and 6E10 (bottom panel), respectively. C. Effects of various y-secretase inhibitors on A� release. Immunoprecipitated A� species were separated on 10% Bicine/urea SDS-PAGE and blotted with 6E10. Ctr stands for control which was treated with DMSO without any y secretase inhibitors. Inhibitors full names were listed in Tab1. 125 � XVI XVII xvm XIX . ii� �i�i�i It) It) ,10¥;'+ ...... At340 ---A.1342 Chapter 4 Fig.3. Effects of inhibitor concentration on AP42/40 ratio. N2aAPPsw/PS 1 wt cells were cultured in the presence of various inhibitors at two doses as indicated in the figure for 1 Oh. The secreted A� spaces were immunoprecipicated from conditioned media and analyzed by 10% Bicine/urea SDS-P AGE system, immunoblotted with6E 10. The bottom panelis the longer exposure of top panel. 126 = = N > � � b � >< .,!, X > X i .,!, � 8 . > >. ..i. -. ai >< = X 1! � > > > > """'>< � < x x � � �� � >< � � -A-AJ3,42fWO 20 2 3 4 5 6 7 8 9 lO H 12 D 14 15 ]6 ...... · .. #/·.. ··� *',;: •.· ·•: · AJ ip.&*.'' ..• �•··. · . --- · ••. �· . · . ..·...... ' �, .. /'• . '"'"'. ;·• .·.*'� �..' . . . . . · · · · . .-...,. . ····-..,.-····.··•·. ·-· W .�. a>--. ----. . ·---. . . ·'--fAPP • · _,.,...... ,_.,....· . � ...... < ...... � -----��...-..-. · · . . .·.•• 1!0l- •• -. - --AJ}46 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Chapter 4 Fig.4. The profile of AP46 and CTFP after removal of y-secretase inhibitors. N2aAPPsw/PS 1 wt cells were cultured in the presence of inhibitors as labeled in the figure for 12 h to pre-accumulated A�46 and CTF�. Conditioned media were removed and cells were washed with cold medium for two times. Fresh media with appropriated y-secretase inhibitors (as labeled in figure, e g. "IV-IV") or without inhibitor (labeled "-" in the figure, eg "IV- ") were added and incubated for 2h before harvesting. Top panel: immunoprecipitated A� species from conditioned media were analyzed by 1 O¾Bicine/urea SOS-PAGE system and blotted with 6E 1 O.Bottom panel: The corresponding lysates were analyzed by the same system as top panel. Note that pre accumulated A�46 did not turnover after removal of compound E (Lanel8, XVIII-) and inhibitor XIX (Lane20) 127 ..c: N � u � b u � Q � ff)- � -ff) I uI I I c::> t .!. � � � · .6 .6 � � ::E ::E � u u u- Q Q :0 � Q Q < ---fAPP I 2 3 4 5 6 7 8 910 ---A �40 �A(342 3 4 5 6 7 8 9 10 Chapter 4 Fig.5. L685,458 and 31C differentially affected Ap46 turnover. N2aAPPsw/PSlwt cells were grown in the absence (lanesl-3) or presence of either DAPM (lanes4-6, 8, and 9) or L685,458(1ane 7) for 12h followed by adding either L685,458(Lanes2,6,9)or 31C( lanes 3 and 8) for 35min pretreatment. Conditioned media were removed and washed with cold fresh media with (Lanes2,6, 7 ,9 with L685,456; lanes 3,8 with 31C; lane 4 with DAPM) or without(lanesland 5) appropriate inhibitors fortwo times (in about 20min). Then freshmedia with or without inhibitors as labeled in each lane were incubated with cells at 3 7 °C for 2h before harvesting conditioned media and cells. LanelO was treated with both L685,458 and Z-V AD from pre-treatment step. Top panel: Lysates were analyzed by 10%Bicine/urea SOS-PAGE system and probed with 6E10. Bottom panel: Immunoprecipicated A� species from each corresponding medium were analyzed by the same system as top panel( DM and L represent DAPM and L685,458, respectively). Note that reduction of A�46 in Lanes 6,8,9 correlates well withincrease on secreted A�. 128 1 2 3 Chapter 5 Fig.1. AfJ46 is present in 20kxg membrane fraction. DAPT treatment and detailed separation precudures were described in the method section. Equal amounts of protein were separated on 10/18% SDS-PAGE, and probed with 6E10 antibody. The lower panel is a part of the longer exposure of the upper one 129 2 3 Chapter 5 Fig.2. The majority of AP46 is resistent to surface trypsinization.Cells surface trypsinized samples were separated on 10/18% SDS-PAGE and probed with 6E10 antibody. "Ctr" and "Tryp" denotes control (without trypsination) and cell surface trypsinized samples, respectively. 130 Monensin C15 Monensin y-adaptin overlap Chapter 5 Fig.3. Treatments of BFA and monensin lead to distribution of APPswe into the ER and TGN in N2a cells, respectively. N2aAPPsw/PS1 wt cells were grown on microscope cover glasses in the presence of absence of 10 µM BF A or 5 µM monensin for 5 h. Then the cells were co-immunostained with appropriate antibodies as described in material and method section. ("Ctr" represents without traffickinginhibitor treatment) 131 Mon(µM) NH4Cl(mM) 10 .Lact 10 2S Tuni Ctr BFA s s Mon(µM) NH4Cl(mM) Bl¼ 5 10 Lact 10 25 Tuni \}:,�,< ·�,;-�,,.;:;; .•-:-:./ :. "J ... ,_ . t' s l 2 3 4 5 6 7 8 a::,... I I I � �- i 2" 4 5 6 8 A 3 7 B Fig 1. effect of or�elle toxins on solubilable APP secretation Chapter 5 Fig.4. The effects of BFA, monensin, lactasystin, NH4CI, and tunicamycin on soluble APP secretion. "Ctr", "BFA", "Mon", "Lact", and "Tuni" represent control, brefeldin A, mo�sin, lactacystin, and tunicamycin, rep!esentively. N2aAPPsw/PS I wt cells were cultured in the absence ( control) or presence of inhibitors as indicated for 8h. A: Lysates were separated on 8/18% two-step SDS-PAGE and probed with 6El0 (top panel), C-1 S(medium panel), and myc (bottompanel) antibodies. B: Conditioned media were separated on 8% SDS-PAGE and probed with either 22Cl I (top panel) or 6El0 (bottom panel) 132 � Co�und E Mon(µM) NH4Cl(mM) Mon . NH4Cl(mM) � Ctr BFA 5 10 Lact 10 25 Tuni c:c. Ctr BFA 5 10 Lact 25 Tuni TFP 1 3 9 11 13 2 4 5 6 7 8 10 12 14 15 16 &, -CTF� -A (346 � 5 7 8 12 13 14 16 A 1 2 3 4 6 9 10 11 JS ,cTFp , ,,n: ---cTFa - '--cTFaend 2 9 10 11 12 13 14 15 16 CTFp ·· -cTFa .. < · .· .· 'CTFaen 9 10 11 12 13 14 15 �AICD 16 AICDend Com�d E � Mon(µM) NH4Cl(mM) Mon( NH4Cl(mM) i Ctr BFA S 10 Lact 10 25 Tuni Ctr BFA S 10 Lact 10 25 Tuni ;�· +if¾+ Af340...... + • P'" Af342.. 1 2 3 ll '12 l3 14 15 16 17 Chapter 5 Fig.5. The effects of BFA, monensin, Iactacystin, NH4Cl, and tunicamycin on y-secretase mediated APP cleavages. N2aAPPsw/PS 1 wt cells were in the presence of inhibitors as indicated for 8 h (lanes9-16 in all with 3nM compound E). A: Lysates were separated either on 10% Bicine/Urea SDS-PAGE (top panel) or on 10/18%SDS PAGE (bottom panel) and probed with 6E 10. B: Lysates were separated on 10/18 SDS PAGE and probed with C15 antibody. The bottom panel is the longer exposure. C: A� species were immunoprecipicated with6E l0, separated on 10% Bicine/urea SDS-PAGE, and probed with 6E 1 0.Lane is synthetic A�40/42peptides 133 LaeZ APPsw-ER APPsw-TCN APPm a!-!o! !o!!a!! .._A ______. =B_l �2--3 _- -4 �5_6_7_8_9______ D fAPP 4. S 6 1 8 9 10 ti ]2 1 2 3 4 S .__E_· ______, --=-F ______. Chapter 5 Fig.6. Addition of the ER sorting signal to APP abolishes both AP40/42 and AfJ46 production whereas addition of the TGN sorting signal did not affect AfJ40/42 and AfJ46 generation. N2aPSlwt cells were transfectedwith lacZ, APPsw-ER, APPsw-TGN, or APPsw vectors ( all in pcDNA3. l), respectively and cutltred in the presence or absence of compound E or L685,456 as indicated for 12h. A: conditioned media (top panel) or lysates (bottom panel) from transiently transfected cells were separated on 8%SDS-PAGE, and probed with 6E10. B: Lysates were separated on 10% Bicine/urea SDS-PAGE and probed with 6El O.C: A� species immunopreciated fromthe conditioned media of transiently transfected cells were separated on 10/18%SDS-PAGE and probed with 6E10. D: Secreted A� species immunopreciated from the conditioned media of transientlytransfected cells were separated on 10% Blcine/urea SDS-PAGE and probed with 6El 0. E: Lysates (top panel) or secreted A� from stable clones as indicated were separated on 10/18% SDS-PAGE and probed with 6E10. F: Secreted A� were separated on 10%Bicine/urea SDS-PAGE andprobed with 6E10 134 APPsw-TGN C 15 APPsw-TGN y-adaptin Overlap Chapter 5 Fig.7. Co-localization of APPsw-ER with Grp78 and APPsw-TGN with y adaptin. N2aAPPsw-ER or N2aAPPsw-TGN Stable cells were cultured on microscope cover glasses and Co-immunuostained with appropriate antibodies as indicated. 135 VITA Guojun Zhao was born in Jiangsu, China in 1963. In 1984, he graduated from Jiangsu Taizhou veterinary school and started to work at the Department of Biochemistry, Jiangsu Agricultural College. He got his master degree in biochemistry from Y angzhou University was appointed as an assistant professor at the same university in 1993. During 1994 and 1995, he did collaborative research at Shanghai Institute of Biochemistry, Chinese Academy of Sciences. From 1996 to 1999, he was a biochemistry lecturer at the College of Bioscience and Biotechnology, Yangzhou University. He came to the Department of Pathology, University of Tennessee, Knoxville as a visiting scholar in November 1999. In 2002 he started to pursue his Ph.D. degree in the major of Comparative and Experimental Medicine at the University of Tennessee, Knoxville. 136 209491 69, 13 ('J 18/17/15 tR3 '