Sie sind auf Seite 1von 8

Neurobiology of Aging 21 (2000) 829 – 836 www.elsevier.

com/locate/neuaging

Familial Alzheimer’s disease presenilin-1 mutants potentiate cell


cycle arrest
Susan M. Janicki, Stacy M. Stabler, Mervyn J. Monteiro*
Medical Biotechnology Center and Department of Neurology, & Division of Human Genetics, University of Maryland, 725 West Lombard Street,
Baltimore, Maryland 21201, USA

Received 18 May 2000; received in revised form 23 August 2000; accepted 28 August 2000

Abstract
We previously reported that overexpression of presenilin 1 and 2 (PS1 and PS2) in HeLa cells leads to cell cycle arrest, and that the
PS2(N141I) FAD mutant potentiates cell cycle arrest compared to wild-type PS2. Using similar BrdU incorporation studies we now report
that three different PS1 FAD mutants also increase cell cycle arrest compared to wild-type PS1when overexpressed in either HeLa cells or
an ATM deficient cell line. We detected reproducible differences in the degrees to which these FAD mutants induced arrest. PS1(P117L)
reduced BrdU incorporation the most (13 to 14%) followed by PS1(P267S) (7.5 to 9%), with the PS1(E280A) mutant inhibiting BrdU
incorporation the least (6 to 7%), compared to wild-type PS1. The degree to which the different mutants inhibited cell cycle progression
correlates somewhat with the age of AD onset induced by the mutations in carriers. Immunoblot analysis of protein extracts from
presenilin-overexpressing cells indicates that the cell cycle-regulated cytoplasmic pool of ␤-catenin is dramatically reduced, whereas the
insoluble ␤-catenin pool remains essentially unaffected. We discuss the implications of these findings in relationship to cell cycle arrest,
apoptosis and AD. © 2000 Elsevier Science Inc. All rights reserved.

Keywords: Presenilins; Alzheimer’s disease; Cell cycle; S phase; Apoptosis; ␤-catenin

1. Introduction somehow interconnected in causing an increase in accumu-


lation of the amyloidgenic A␤ fragment [8].
Alzheimer’s disease (AD), the most common cause of PS1 and PS2 are multi-pass trans-membrane proteins
dementia in the elderly, is characterized by progressive loss which share 67% sequence identity, however, there is de-
of memory and a decline in cognitive function. The mech- bate regarding their precise topology. A commonly drawn
anisms that cause AD pathogenesis are unknown although model depicts a polypeptide with eight trans-membrane
there is strong evidence that the disease is caused by the domains, such that the N- and C-termini and a large “loop”
interplay of several factors including age, genes, and the between trans-membrane domains 6 and 7 are all oriented
environment. The majority of AD cases are late-onset, ap- toward the cytoplasm [9]. Although PS1 and PS2 share
pearing in people over that age of 65 years, however, a extensive sequence identity along their entire lengths, their
small percentage of cases (⬃5%), are early onset with N-terminal domains and the second half of their loops are
inception of disease occurring between the ages of 20 to 65 highly divergent, suggesting that these unique regions could
years. Many of the early onset cases have been traced to modulate potentially different functions of the two preseni-
familial autosomal dominant inheritance of mutations in lins. Examination of presenilin-linked familial AD (FAD)
three genes: the ␤-amyloid precursor protein (APP) and two mutations reveals that most are missense mutations in res-
homologous genes, presenilin-1 and 2 (PS1 and PS2) [8]. idues conserved between the two polypeptides, suggesting
The precise mechanisms by which mutations in these three that these conserved residues are important for proper func-
genes cause AD are not known, although they all appear tion.
Studies of presenilin homologs in a number of species
have indicated that the presenilin genes are required for
* Corresponding author. Tel.: ⫹1-410-706-8132; fax: ⫹1-410-706-1732. proper development. Mutation of the Caenorhabditis el-
E-mail address: monteiro@umbi.umd.edu (M.J. Monteiro) egans presenilin gene, sel-12, produce defects in cell sig-

0197-4580/00/$ – see front matter © 2000 Elsevier Science Inc. All rights reserved.
PII: S 0 1 9 7 - 4 5 8 0 ( 0 0 ) 0 0 2 2 2 - 0
830 S.M. Janicki et al. / Neurobiology of Aging 21 (2000) 829 – 836

naling involving Notch-based receptors [9,16]. Disruption 2.3. Immunofluorescence and BrdU staining
of the Drosophila melanogaster presenilin gene is an em-
bryonic lethal with defective embryos displaying severe Exponentially growing cells were plated on glass cover-
Notch-like phenotypes [31,40]. Mammalian presenilins also slips, and transfected with plasmid expression constructs
play important roles in early development, as disruption of using the CaPO4 coprecipitation method [12]. BrdU was
the mouse PS1 gene leads to death shortly after birth with added to the media immediately after the transfection. At
embryos displaying central nervous system defects together 19h post-transfection the cells were fixed in 50 mM Gly-
with abnormal patterning of the axial skeleton and spinal cine/70% ethanol at ⫺20°C for 20 min, rehydrated in PBS,
ganglia [28,39]. Interestingly, PS1⫺/⫺ mice can be rescued blocked in 0.8% BSA, and stained for expressed proteins.
by human transgenes containing FAD-linked mutations, in- Cells were next stained for BrdU incorporation (5-Bromo-
dicating that the FAD mutations do not affect presenilin 2⬘-deoxy-uridine Labeling and Detection Kit I, Boehringer
functions related to embryo development [4,26]. In contrast, Mannheim), and with 1 ␮g/ml 4⬘6-diamidino-2-phenylin-
disruption of the mouse PS2 gene produces no obvious dole (DAPI) as previously described [13].
defects, but PS1/PS2 double knock-out mice die earlier at Antibodies used for the immunofluorescence staining of
embryonic day 9.5 and, like PS1⫺/⫺ mice, display severe PS proteins were described previously [13]. The stained
misexpression of proteins involved in Notch signaling coverslips were viewed under a Leica DRB microscope
[6,10]. These results indicate that there is functional redun- using a double FITC/Rhodamine cube to simultaneously
dancy between PS1 and PS2; PS1 can compensate for PS2, view PS staining (Rhodamine) and BrdU staining (FITC).
but PS2 cannot compensate for PS1, at least during early By scanning the slides with the double FITC/Rhodamine
mouse development. cube the percentage of PS transfected cells that stained
A number of studies have indicated that presenilins are positive for BrdU was determined. At least three popula-
involved in the regulation of apoptosis, and that cells ex- tions of cells on independent coverslips were counted for
pressing AD associated mutations in either PS1 or PS2 die each construct evaluated. P values were calculated by com-
faster than cells expressing wild-type presenilins [3,5,7,37, paring means in an unpaired two-tailed t test with the
38,42]. Consistent with these findings we previously Statview 512 program (Brain Power Inc.).
showed that when overexpressed in epithelial HeLa cells the
PS2(N141I) FAD mutant induces increased apoptosis com- 2.4. Fractionation of cells into soluble and insoluble
pared to wild-type PS2 [12]. By BrdU incorporation studies protein pools
we further demonstrated that presenilin overexpression in
exponentially growing cultures leads to arrest in the G1 HeLa cells transiently transfected with presenilin con-
phase of the cell cycle, and that the PS2(N141I) mutant structs were fractionated into soluble and insoluble protein
potentiates the cell cycle arrest [13]. In this report we pools [19]. Dishes of transfected cells were washed two
demonstrate that cell cycle arrest is also potentiated by PS1 times in ice-cold PBS, and the cells were scraped from the
FAD mutants and that ␤-catenin levels are altered upon dishes in 600 ␮l of PBS, spun down at 15,000 g for two
presenilin overexpression. minutes (4°C), and resuspended in 150 ␮l of TEA buffer 1
(10 mM Triethanolamine-HCl (TEA), pH 7.4, 10 mM KCl,
1.5 mM MgCl2, 5 mM Iodoacetamide, and 1 mM Pefabloc).
2. Materials and methods Cells were left on ice for 10 min and then homogenized
15–20 times. 150 ␮l of TEA buffer 2 (10 mM TEA pH 7.4,
2.1. Construction of PS expression plasmids 270 mM KCl, 1.5 mM MgCl2, 5 mM Iodoacetamide, and 1
mM Pefabloc) was then added, and the cells were vortexed
The cloning of wild-type PS1 and PS2, and the FAD and spun at 100,000 g for 45 min in the ultracentrifuge at
PS2(N141I) mutant into a CMV expression plasmid was 4°C. The supernatants were removed from the pellets, and
previously described [12,13]. The wild-type PS1 expression the pellets were suspended in 300 ␮l of lysis buffer and
construct was used to prepare mutants corresponding to sonicated. BCA protein assays were used to measure the
FAD PS1(P267S), PS1(E280A) and PS1(P117L) mutations protein concentration of each fraction, and equal amounts of
by PCR modification using appropriately designed primers protein from each fraction were run on 8.5% SDS-PAGE
as described previously [30]. gels and transferred to nitrocellulose membranes.

2.2. Cell lines, cell culture 2.5. Protein preparation, SDS electrophoresis, and
immunoblotting
HeLa cells were grown in DMEM supplemented with
10% FBS [20], and an ataxia-telangiectasia (AT) trans- Cells were lysed in a buffer containing protease inhibi-
formed cell line (GMO9607A, NIGMS Human Genetic tors [20], and protein concentration was determined by the
Mutant Cell Repository) was grown in MEM supplemented BCA protein assay (Pierce). For immunoblot analysis of PS
with 10% FBS. expression, lysates were mixed with Laemmli buffer con-
S.M. Janicki et al. / Neurobiology of Aging 21 (2000) 829 – 836 831

taining 8 M urea and 15 mM DTT and electrophoresed


unboiled on 8.5% SDS-polyacrylamide gels. Lysates used
for analysis of ␤-catenin and ␤-tubulin were boiled in Lae-
mmli buffer and electrophoresed on 8.5% gels. The frac-
tionated proteins were transferred onto 0.2 ␮m nitrocellu-
lose membranes by electroblotting and processed as
described previously [12]. Antibodies used were the follow-
ing: a goat N-terminal specific anti-PS1 antibody (Santa
Cruz Biotechnology), a goat anti-␤-catenin antibody (Santa
Cruz Biotechnology) and a mouse monoclonal anti-␤-tubu-
lin antibody (Sigma).

3. Results

3.1. Transient overexpression of PS1 blocks cell cycle


progression: PS1 FAD mutants potentiate cell cycle
arrest

Our previous studies had indicated that the overexpres-


sion of presenilin 1 and 2 genes in HeLa and ATM⫺/⫺
cells induced cell cycle arrest [13]. Moreover, we demon-
strated by BrdU incorporation studies that cell cycle pro-
gression was more inhibited by PS2 proteins containing the
PS2(N141I) FAD mutation compared to the wild-type pro-
tein. In this study we investigated whether overexpression
of PS1 FAD mutants also potentiates cell cycle arrest. Three
different PS1 mutants PS1(P267S), PS1(E280A) and Fig. 1. Immunoblot analysis of PS1 transfected HeLa cell lysates. Equiv-
PS1(P117L), which cause early onset AD (1,11,15,36; http: alent amounts of protein lysates from PS1 transfected cells immunoblotted
with a PS1 specific antibody. Lysates were from cells transfected with the
molgen-www.uia.ac.be/ADMutations/) were studied. Of the following constructs: Lane 1, mock-transfected (control); lane 2, wild-type
three mutations, the PS1(P117L) mutation is one of the PS1; lane 3, PS1(P267S); lane 4, PS1(E280A); and lane 5, PS1(P117L).
most aggressive FAD mutations described so far, causing Full-length uncleaved PS1 is indicated by the arrow.
disease as early as the third decade of life [36]. Transient
transfection of wild-type PS1 and the three PS1 FAD mu-
tants in HeLa cells followed by immunoblotting of protein ingly, however, was that BrdU incorporation was reduced to
lysates indicated that all constructs were expressed at sim- a greater extent upon overexpression of the three PS1 FAD
ilar levels (Fig. 1). The full-length PS1 polypeptides mi- mutants compared to the wild-type protein (Fig. 3). BrdU
grated as an approximately 45 kDa band (Fig. 1, lanes 2–5; labeling was reduced 6.7% ((46.9 –53.1)/92.3 ⫻ 100) for the
indicated by an arrow) and several high molecular weight PS1(E280A) mutant, 9.5% for the PS1(P267S) mutant, and
bands, the latter probably corresponding to PS1 complexes. 13.0% for the PS1(P117L) mutant compared to BrdU label-
As expected these bands were not detected in untransfected ing obtained with wild-type PS1. Collectively, these results
HeLa cells (Fig. 1, Lane 1). indicate that the PS1 FAD mutants potentiate cell cycle
To determine the effects of overexpressing PS1 on cell arrest in exponentially growing HeLa cells.
cycle progression, HeLa cells exponentially growing on Previously we had demonstrated that cell cycle arrest due
coverslips were transfected with equivalent amounts of to overexpression of presenilins in dividing cells also man-
plasmid DNA encoding either wild-type or mutant PS1 ifests in ATM⫺/⫺ cells, which lack functional ATM pro-
proteins. Immediately following the transfection, BrdU was tein, a kinase induced by double-stranded DNA breaks and
added to the medium and at 19 h the coverslips were fixed which activates cell cycle check-point [13,34]. To deter-
and double stained for PS1 and BrdU [13]. The stained mine if ATM protein kinase is involved in presenilin in-
coverslips were then examined by immunofluorescence mi- duced cell cycle arrest we studied the consequence of over-
croscopy and the percentage of PS1 expressing cells that expression of PS1 FAD mutants in the ATM⫺/⫺ cell line.
had incorporated BrdU into their genomes was determined. Fig. 4 demonstrates that BrdU labeling was reduced 6.1%
As shown in Figs. 2 and 3 overexpression of wild-type PS1 ((36.4 – 41.8)/87.5 ⫻ 100) for the PS1(E280A) mutant,
significantly reduced BrdU labeling (53.1% ⫾ 3.2) com- 7.5% for the PS1(P267S) mutant, and 14.2% for the
pared mock-transfected control cells (92.3% ⫾ 3.0), which PS1(P117L) mutant in comparison to wild-type PS1 in the
is consistent with our previous findings [13]. More interest- ATM⫺/⫺ cell line. When taken together these results dem-
832 S.M. Janicki et al. / Neurobiology of Aging 21 (2000) 829 – 836

Fig. 2. Overexpression of PS1 in HeLa cells inhibits cell cycle progression. HeLa cells were mock-transfected (a-c) or transiently transfected with PS1 (d-f)
and evaluated for BrdU incorporation. Left panels show transfected protein, middle panels are BrdU labeling, and right panels are DAPI DNA images. Only
few of the PS overexpressing cells replicated their DNA as revealed by BrdU labeling while non-PS1 overexpressing cells were not inhibited in DNA
replication. Bar: 10 ␮m.

onstrate that PS1 FAD mutants cause increased cell cycle FAD mutants, PS1(P117L) and PS2(N141I). Although we
arrest when overexpressed in two different cell lines and were unable to detect any significant difference in the levels
that ATM protein kinase is not necessary for inducing of ␤-catenin when we compared whole cell lysates of un-
arrest. transfected and presenilin-transfected cells (data not
shown), when soluble and insoluble protein fractions were
3.2. Overexpression of presenilins proteins in HeLa cells evaluated, cytoplasmic ␤-catenin levels were significantly
reduces cytoplasmic ␤-catenin levels reduced (40 to 60%) in the soluble fraction of presenilin-
transfected cells compared to mock-transfected control cells
␤-catenin is a 92–97 kDa protein that is present in cells (Fig. 5, lanes 1–5). In contrast, there was not a dramatic
in two forms, cytoplasmic (soluble) and a membrane-asso- difference between the levels of ␤-catenin in the insoluble
ciated form that have separate functions; signaling and ad- fraction (Fig. 5, lanes 6 –10). We confirmed that equivalent
hesion, respectively. As ␤-catenin has been shown to inter- amounts of protein were loaded in the two sets of fractions
act with the presenilins [21,23,41,44], and because by evaluating ␤-tubulin immunoreactivity on the same blot;
cytoplasmic ␤-catenin levels oscillate during the cell cycle the levels appeared to be approximately the same in all of
[25], increasing dramatically during the G1 to S phase the lanes (Fig. 5).
transition, we investigated whether presenilin-induced cell Although all four presenilin constructs had reduced lev-
cycle arrest is correlated with changes in ␤-catenin accu- els of soluble ␤-catenin compared to the mock-transfected
mulation. Thus wild-type and mutant presenilin constructs cells (Fig. 5, lanes 1–5) we were unable to detect any
were overexpressed in HeLa cells and the level of ␤-catenin reproducible difference between the wild-type and corre-
accumulation were evaluated by immunoblot analysis. Im- sponding FAD presenilin mutants using this fractionation
munofluorescence staining of the transiently transfected assay. We believe that this due to technical difficulties
cells indicated that approximately 60 –70% of the cells were associated in measuring small differences (presumably a
expressing presenilin proteins, similar to the transfection 6 –13% difference based on the cell cycle inhibition data)
efficiencies obtained previously [13]. The level of ␤-catenin between wild-type and mutant proteins by immunoblotting
in mock-transfected HeLa cells were compared to those in of the transiently transfected lysates. The ␤-catenin immu-
cells over-expressing wild-type PS1 and PS2 and the two noreactivity was detected as a doublet band (approximately
S.M. Janicki et al. / Neurobiology of Aging 21 (2000) 829 – 836 833

Fig. 3. Quantification of BrdU incorporation of exponentially growing Fig. 4. Quantification of BrdU incorporation of exponentially growing
HeLa cells transfected with wild-type and PS1 FAD mutants. (A) Wild- ATM deficient cells transfected with wild-type and PS1 FAD mutants. (A)
type PS1 and three FAD mutants, PS1(P267S), PS1(E280A), and Wild-type PS1 and three FAD mutants, PS1(P267S), PS1(E280A), and
PS1(P117L), were transiently transfected into exponentially growing HeLa PS1(P117L), were transiently transfected into exponentially growing HeLa
cells. BrdU was added to the media at the time of transfection, and cells cells. BrdU was added to the media at the time of transfection, and cells
were fixed 19 h post-transfection. Compared to mock-transfected control were fixed 19 h post-transfection. Compared to mock-transfected control
cells (92.3% ⫾ 3.0, n ⫽ 3), BrdU labeling was significantly reduced by cells (87.5% ⫾ 0.9, n ⫽ 3), BrdU labeling was significantly reduced by
PS1 (53.1% ⫾ 3.2, n ⫽ 8; p value ⬍ 0.0001). When compared to wild-type PS1 (41.8% ⫾ 0.8, n ⫽ 6). When compared to wild-type PS1 the three
PS1 the three FAD mutants significantly reduced BrdU incorporation; FAD mutants significantly reduced BrdU incorporation; PS1(P267S)
PS1(P267S) (44.3% ⫾ 1.4, n ⫽ 3), PS1(E280A) (47.0% ⫾ 0.9, n ⫽ 3) and (35.2% ⫾ 3.6, n ⫽ 3), PS1(E280A) (36.4% ⫾ 1.7, n ⫽ 6) and PS1(P117L)
PS1(P117L) (41.1% ⫾ 1.6, n ⫽ 8), p values of 0.0016; 0.0111, and 0.0001, (29.3% ⫾ 5.6, n ⫽ 4), p values of 0.0027; 0.0001, and 0.0006,
respectively). respectively).

92–97 kDa) in the soluble fraction whereas only one band when overexpressed in HeLa cells [12,13]. Here we dem-
was seen in the insoluble fraction. The higher of the two onstrate that like PS2, three different PS1 FAD mutants also
␤-catenin bands probably corresponds to ubiquitinated increase cell cycle arrest as assayed by reduced incorpora-
␤-catenin [24]. These results suggest that overexpression of tion of BrdU in transfected HeLa cells. Potentiation of the
presenilins correlates with a reduction in the soluble pool of cell cycle arrest by the PS1 FAD mutants was also observed
␤-catenin which is involved in signaling. in ATM⫺/⫺ cells, indicating that the effects are indepen-
dent of ATM protein, a kinase that is involved in cell cycle
check-point control and which is activated in response to
4. Discussion double-stranded DNA breaks [34]. The increased cell cycle
arrest potentiated by FAD mutants may represent a funda-
Previously we demonstrated that, when compared to the mental mechanism by which the mutant proteins act in cells.
wild-type protein, PS2 proteins containing the N141I FAD Interestingly the three FAD mutants inhibited BrdU labeling
mutation cause increased cell cycle arrest and apoptosis to different extents. Despite inherent experimental errors in
834 S.M. Janicki et al. / Neurobiology of Aging 21 (2000) 829 – 836

Fig. 5. Immunoblot analysis of ␤-catenin levels in soluble and insoluble fractions of presenilin transfected cells. Equal protein amounts of soluble (lanes 1–5)
and insoluble fractions (lanes 6 –10) prepared from mock-transfected (lane 1 and 6, respectively) and PS transfected (lanes 2–5 and 7–10, respectively) were
immunoblotted with ␤-catenin and ␤-tubulin antibodies. Presenilin overexpression decreases the amount of ␤-catenin in the soluble cytoplasmic pool but does
not affect ␤-catenin levels in the insoluble fraction, whereas ␤-tubulin remained relatively unchanged. The band intensities of ␤-catenin and tubulin in PS
transfected cells were determined and expressed as a ratio of ␤-catenin to tubulin after normalization to the ratio determined for mock transfected cells:
Soluble fraction, Control mock-transfected ⫽ 1.0, PS1 ⫽ 0.57, PS1(P117L) ⫽ 0.48, PS2 ⫽ 0.45, PS2(N141I) ⫽ 0.40; Pellet fraction, Control mock-
transfected ⫽ 1.0, PS1 ⫽ 0.85, PS1(P117L) ⫽ 0.85, PS2 ⫽ 1.0, PS2(N141I) ⫽ 0.94.

normalizing expression levels between different constructs needed to determine whether all FAD mutations affect cell
in transiently transfected cells, we found reproducible dif- cycle progression and whether cell cycle changes in divid-
ferences by which each of the mutants inhibited BrdU ing cells can be extrapolated to loss of cells, including
incorporation when assayed in multiple experiments and terminally differentiated neurons in AD.
utilizing two different cell lines. Of the three PS1 mutants Our results showing presenilin overexpression halts the
studied, the PS1(P117L) reduced BrdU incorporation the cell cycle may be physiologically relevant to the normal
most (13 to 14%) followed by PS1(P267S) (7.5 to 9%), and function of presenilins in dividing cells. Recent evidence in
the PS1(E280A) mutant inhibiting BrdU incorporation the transgenic mice has indicated that presenilins are involved
least (6 to 7%). It is noteworthy that the different degree by in regulating cellular proliferation of keratinocytes. Thus,
which the three mutants potentiated arrest correlates with mice disrupted of the murine presenilin 1 (PS1) gene die at
the approximate age of AD onset in patients carrying these day 1 after birth, but are rescued and survive to adulthood
mutations [32]. PS1(P117L) which inhibited BrdU labeling upon expression of human PS1 transgene under the control
the most is one of the most aggressive FAD mutations of the Thy-1 promoter [26]. However, these rescued mice
described, causing disease at early as 27–28 years of age develop extensive epidermal hyperplasia and neoplasia
[36]. The two other mutations PS1(P267S) and PS1(E280A) [43]. The abnormalities appear to arise due to the restricted
have been documented to cause disease at a slightly later expression of the Thy-1 driven PS1 transgene to neurons
age, 35 and 45–54 years, respectively [32]. Thus it is con- and its failure, unlike the normal murine PS1 gene, to be
ceivable that the degree of cell cycle inhibition by the expressed in skin. These results are consistent with our
different PS1 FAD mutants and the age of AD onset may be findings, that increased expression of presenilins halts the
related. However, since our studies are based on only a cell cycle, and conversely we presume loss of presenilin
small number of mutants, and since we do not know what expression leads to cellular proliferation. In may be fortu-
role these cell cycle changes play in AD pathogenesis, it is itous that the presenilin effects on the cell cycle we describe
not possible at this time to conclude with any certainty were obtained utilizing HeLa cells, an epithelial derived cell
whether the two processes are linked. Interestingly, Li et al. line.
[17] localized presenilins to centrosomes and centromeres In an effort to determine the mechanism by which pre-
in fibroblasts, and have proposed that presenilins are in- senilin overexpression induces cell cycle arrest, we used
volved in chromosome segregation, and that FAD presenilin immunoblot analysis to determine if the expression profile
mutations cause increased chromosome missegregation of a number of known cell cycle regulated proteins were
leading to inappropriate apoptosis. Further studies will be altered. As documented previously, presenilin overexpres-
S.M. Janicki et al. / Neurobiology of Aging 21 (2000) 829 – 836 835

sion in HeLa cells leads to arrest of cells in the G1 phase of Acknowledgments


the cell cycle. However, despite arrest at this phase of the
cell cycle, we did not detect any noticeable changes in the We thank Dr. Ann Pluta for critical comments on the
expression patterns of p21, 27, pRb, c-myc, and p53 pro- manuscript. Supported by NIA grant AG16839 to MJM.
teins in cell lysates of untransfected and presenilin overex-
pressing cells [13]. Here we demonstrate that the overex-
pression of wild-type and mutant presenilin proteins causes References
a specific and marked decrease in cytoplasmic (soluble)
pool of ␤-catenin. The decrease in the cytoplasmic pool of [1] Alzheimer’s Disease Collaborative Group. The structure of the pre-
␤-catenin upon presenilin overexpression could be respon- senilin 1 (S182) gene and identification of six novel mutations in
sible for the cell cycle arrest, since ␤-catenin plays an early onset AD families. Nat Genet 1995;11:219 –22.
[2] Busser J, Geldmacher DS, Herrup K. Ectopic cell cycle proteins
important role in cell cycle progression [25]. In an elegant predict the sites of neuronal cell death in Alzheimer’s disease brain.
study, Orford et al. [25] demonstrated that the cytoplasmic J Neurosci 1998;18:2801–7.
pool of ␤-catenin oscillates, increasing rapidly between G1 [3] Cotman CW. Apoptosis decision cascades and neuronal degeneration
and S phase of the cell cycle. Furthermore, they demon- in Alzheimer’s disease. Neurobiol Aging 1998;19:S29 –S32.
strated that the elevation in cytoplasmic ␤-catenin is impor- [4] Davis JA, Naruse S, Chen H, Eckman C, Younkin S, Price DL,
Borchelt DR, Sisodia SS, Wong PC. An Alzheimer’s disease-linked
tant for driving cells from G1 into S phase. Interestingly,
PS1 variant rescues the developmental abnormalities of PS1-deficient
they also showed that ␤-catenin can function as a potent embryos. Neuron 1998;20:603–9.
inhibitor of apoptosis since transgenic expression of ␤-cate- [5] Deng G, Pike CJ, Cotman CW. Alzheimer-associated presenilin-2
nin inhibited apoptosis in their model system. Based on their confers increased sensitivity to apoptosis in PC12 cells. FEBS Lett
results they proposed that loss of cytoplasmic ␤-catenin 1996;397:50 – 4.
leads to cell cycle arrest at G1/S phase and under certain [6] Donoviel DB, Hadjantonakis AK, Ikeda M, Zheng H, Hyslop PS,
Bernstein A. Mice lacking both presenilin genes exhibit early embry-
circumstances this leads to apoptosis. Our findings are con-
onic patterning defects. Genes Dev 1999;13:2801–10.
sistent with this hypothesis, as we have shown presenilin [7] Guo Q, Sopher BL, Furukawa K, Pham DG, Robinson N, Martin GM,
expression in HeLa cells causes cell cycle arrest at G1 and Mattson MP. Alzheimer’s presenilin mutation sensitizes neural cells
increased apoptosis [12,13]. The decrease in ␤-catenin due to apoptosis induced by trophic factor withdrawal and amyloid beta-
to presenilin overexpression may prevent cells from reach- peptide: involvement of calcium and oxyradicals. J Neurosci 1997;
ing a critical threshold level of ␤-catenin that is necessary 11:4212–22.
[8] Hardy J. Amyloid, the presenilins and Alzheimer’s disease. Trends
for progressing into S phase, leading to G1 arrest that
Neurosci 1997;4:154 –9.
culminates in apoptosis. [9] Haass C, De Strooper B. The presenilins in Alzheimer’s disease–
There is considerable support for the idea that presenilins proteolysis holds the key. Science 1999;286:916 –9.
may be involved in modulating ␤-catenin stability [14,21, [10] Herreman A, Hartmann D, Annaert W, Saftig P, Craessaerts K,
23,35,42]. Due to technical reasons we were unable to Serneels L, Umans L, Schrijvers V, Checler F, Vanderstichele H,
detect any reproducible differences in ␤-catenin levels in Baekelandt V, Dressel R, Cupers P, Huylebroeck D, Zwijsen A, Van
Leuven F, De Strooper B. Presenilin-2 deficiency causes a mild
cells overexpressing wild-type and FAD presenilin proteins.
pulmonary phenotype and no changes in amyloid precursor protein
However, Zhang et al. [42] and Weihl et al. [35] have processing but enhances the embryonic lethal phenotype of preseni-
provided strong evidence that PS1 FAD mutations destabi- lin-1 deficiency. Proc Natl Acad Sci USA 1999;96:11872–7.
lize ␤-catenin to a greater extent than the wild-type protein, [11] Hutton M, Busfield F, Wragg M, Crook R, Perez-Tur J, Clark RF,
and that this destabilization potentiates apoptosis in trans- Prihar G, Talbot C, Phillips H, Wright K, Baker M, Lendon C, Duff
fected neuronal cells. K, Martinez A, Houlden H, Nichols A, Karran E, Roberts G, Roques
P, Rossor M, Venter JC, Adams MD, Cline RT, Phillips CA, Goate
Our observations that FAD mutations in presenilin 1 and
A. Complete analysis of the presenilin 1 gene in early onset Alzhei-
2 genes compromise cell cycle progression in dividing cells mer’s disease. Neuroreport 1996;7:801–5.
needs to be integrated with the now well-document evi- [12] Janicki S, Monteiro MJ. Increased apoptosis arising from increased
dence for misexpression of cell cycle proteins in AD af- expression of the Alzheimer’s disease-associated presenilin-2 muta-
fected brains [2,18,22,27,29,33]. As discussed previously tion (N141I). J Cell Biol 1997;139:485–95.
[13], the misregulation of cell cycle proteins in AD may [13] Janicki SM, Monteiro MJ. Presenilin overexpression arrests cells in
the G1 phase of the cell cycle. Arrest potentiated by the Alzheimer’s
reflect an abortive attempt of the cells to reenter the cell
disease PS2(N141I) mutant. Am J Pathol 1999;155:135– 44.
cycle, or it could represent a protective response of the cells [14] Kang DE, Soriano S, Frosch MP, Collins T, Naruse S, Sisodia SS,
to death stimuli. With regard to these possibilities, it will be Leibowitz G, Levine F, Koo EH. Presenilin 1 facilitates the consti-
important to determine whether autosomal dominant muta- tutive turnover of beta-catenin: differential activity of Alzheimer’s
tions in presenilin genes cause increased apoptosis in the disease-linked PS1 mutants in the beta-catenin-signaling pathway.
brains of affected individuals and whether cell cycle pro- J Neurosci 1999;19:4229 –37.
[15] Lendon CL, Martinez A, Behrens IM, Kosik KS, Madrigal L, Norton
teins are misexpressed in these individuals. A better under-
J, Neuman R, Myers A, Busfield F, Wragg M, Arcos M, Arango
standing of the mechanisms that lead to the upregulation of Viana JC, Ossa J, Ruiz A, Goate AM, Lopera F. E280A PS-1
these cell cycle proteins may eventually lead to better in- mutation causes Alzheimer’s disease but age of onset is not modified
tervention strategies in the effort to prevent and/or cure AD. by ApoE alleles. Hum Mutat 1997;10:186 –95.
836 S.M. Janicki et al. / Neurobiology of Aging 21 (2000) 829 – 836

[16] Levitan D, Greenwald I. Facilitation of lin-12-mediated signalling by [30] Stabler SM, Ostrowski LL, Janicki SM, Monteiro MJ. A myristoy-
sel-12, a Caenorhabditis elegans S182 Alzheimer’s disease gene. lated calcium-binding protein that preferentially interacts with the
Nature 1995;377:351– 4. Alzheimer’s disease presenilin-2 protein. J Cell Biol 1999;145:1277–
[17] Li J, Xu M, Zhou H, Ma J, Potter H. Alzheimer presenilins in the 92.
nuclear membrane, interphase kinetochores, and centrosomes suggest [31] Struhl G, Greenwald I. Presenilin is required for activity and nuclear
a role in chromosome segregation. Cell 1997;90:917–27. access of Notch in Drosophila. Nature 1999;398:522–5.
[18] McShea A, Harris PL, Webster KR, Wahl AF, Smith MA. Abnormal [32] Van Broeckhoven C. Presenilins, and Alzheimer disease. Nat Genet
expression of the cell cycle regulators P16 and CDK4 in Alzheimer’s 1995;11:230 –2.
disease. Am J Pathol 1997;150:1933–9. [33] Vincent I, Jicha G, Rosado M, Dickson DW. Aberrant expression of
[19] Mical TI, Monteiro MJ. The role of sequences unique to nuclear mitotic cdc2/cyclin B1 kinase in degenerating neurons of Alzheimer’s
intermediate filaments in the targeting and assembly of human lamin
disease brain. J Neurosci 1997;17:3588 –98.
B: evidence for lack of interaction of lamin B with its putative
[34] Wang JYJ. New link in a web of human genes. Nature 2000;405:
receptor. J Cell Sci 1998;111:3471– 85.
404 –5.
[20] Monteiro MJ, Mical TI. Resolution of kinase activities during the
[35] Weihl CC, Miller RJ, Roos RP. The role of beta-catenin stability in
HeLa cell cycle: identification of kinases with cyclic activities. Exp
mutant PS1-associated apoptosis. Neuroreport 1999;10:2527–32.
Cell Res 1996;223:443–51.
[21] Murayama M, Tanaka S, Palacino J, Murayama O, Honda T, Sun X, [36] Wisniewski T, Dowjat WK, Buxbaum JD, Khorkova O, Efthimio-
Yasutake K, Nihonmatsu N, Wolozin B, Takashima A. Direct asso- poulos S, Kulczycki J, Lojkowska W, Wegiel J, Wisniewski HM,
ciation of presenilin-1 with beta-catenin. FEBS Letters 1998;433: Frangione B. A novel Polish presenilin-1 mutation (P117L) is asso-
73–7. ciated with familial Alzheimer’s disease and leads to death as early as
[22] Nagy Z, Esiri MM, Cato AM, Smith AD. Cell cycle markers in the the age of 28 years. Neuroreport 1998;9:217–2.
hippocampus in Alzheimer’s disease. Acta Neuropathol 1997;94:6 – [37] Wolozin B, Iwasaki K, Vito P, Ganjei JK, Lacana E, Sunderland T,
15. Zhao B, Kusiak JW, Wasco W, D’Adamio L. Participation of prese-
[23] Nishimura M, Yu G, Levesque G, Zhang DM, Ruel L, Chen F, nilin 2 in apoptosis: enhanced basal activity conferred by an Alzhei-
Milman P, Holmes E, Liang Y, Kawarai T, Jo E, Supala A, Rogaeva mer mutation. Science 1996;274:1710 –3.
E, Xu DM, Janus C, Levesque L, Bi Q, Duthie M, Rozmahel R, [38] Wolozin B, Alexander P, Palacino J. Regulation of apoptosis by
Mattila K, Lannfelt L, Westaway D, Mount HT, Woodgett J, St presenilin 1. Neurobiol Aging 1998;19:S23–S27.
George-Hyslop P. Presenilin mutations associated with Alzheimer [39] Wong PC, Zheng H, Chen H, Becher MW, Sirinathsinghji DJ, Trum-
disease cause defective intracellular trafficking of beta-catenin, a bauer ME, Chen HY, Price DL, Van der Ploeg LH, Sisodia SS.
component of the presenilin protein complex. Nat Med Presenilin 1 is required for Notch1 and DII1 expression in the parax-
1999;5:164 –9. ial mesoderm. Nature 1998;387:288 –92.
[24] Orford K, Crockett C, Jensen JP, Weissman AM, Byers SW. Serine [40] Ye Y, Lukinova N, Fortini ME. Neurogenic phenotypes and altered
phosphorylation-regulated ubiquitination and degradation of beta- Notch processing in Drosophila Presenilin mutants. Nature 1999;398:
catenin. J Biol Chem 1997;272:24735– 8. 525–9.
[25] Orford K, Orford CC, Byers SW. Exogenous expression of beta- [41] Yu G, Chen F, Levesque G, Nishimura M, Zhang DM, Levesque L,
catenin regulates contact inhibition, anchorage-independent growth,
Rogaeva E, Xu D, Liang Y, Duthie M, St George-Hyslop PH, Fraser
anoikis, and radiation-induced cell cycle arrest. J Cell Biol 1999;146:
PE. The presenilin 1 protein is a component of a high molecular
855– 68.
weight intracellular complex that contains beta-catenin. J Biol Chem
[26] Qian S, Jiang P, Guan XM, Singh G, Trumbauer ME, Yu H, Chen
1998;273(26):16470 –16475.
HY, Van de Ploeg LH, Zheng H. Mutant human presenilin 1 protects
[42] Zhang Z, Hartmann H, Do VM, Abramowski D, Sturchler-Pierrat C,
presenilin 1 null mouse against embryonic lethality and elevates
Abeta1– 42/43 expression. Neuron 1998;20:611–7. Staufenbiel M, Sommer B, van de Wetering M, Clevers H, Saftig P,
[27] Raina AK, Monteiro MJ, McShea A, Smith MA. The role of cell De Strooper B, He X, Yankner BA. Destabilization of beta-catenin by
cycle-mediated events in Alzheimer’s disease. Int J Exp Pathol 1999; mutations in presenilin-1 potentiates neuronal apoptosis. Nature
80:71– 6. 1998;395:698 –702.
[28] Shen J, Bronson RT, Chen DF, Xia W, Selkoe DJ, Tonegawa S. [43] Zheng H, Xia X-F, Wu Y, Wu X. Loss of presenilin 1 leads to
Skeletal, and CNS defects in Presenilin-1-deficient mice. Cell 1997; epidermal hyperplasia and tumorgenesis in adult mice. Neurobiol
89:629 –39. Aging 2000;21:S133.
[29] Smith MZ, Nagy Z, Esiri MM. Cell cycle-related protein expression [44] Zhou J, Liyanage U, Medina M, Ho C, Simmons AD, Lovett M,
in vascular dementia and Alzheimer’s disease. Neurosci Lett 1999; Kosik KS. Presenilin 1 interaction in the brain with a novel member
271:45– 8. of the Armadillo family. Neuroreport 1997;8:1489 –94.

Das könnte Ihnen auch gefallen