Presenilin/gamma-secretase-dependent processing of beta-amyloid precursor protein regulates EGF receptor expression.
ABSTRACT Presenilins (PS, PS1/PS2) are necessary for the proteolytic activity of gamma-secretase, which cleaves multiple type I transmembrane proteins including Alzheimer's beta-amyloid precursor protein (APP), Notch, ErbB4, etc. Cleavage by PS/gamma-secretase releases the intracellular domain (ICD) of its substrates. Notch ICD translocates into the nucleus to regulate expression of genes important for development. However, the patho/physiological role of other ICDs, especially APP ICD (AICD), in regulating gene expression remains controversial because evidence supporting this functionality stems mainly from studies performed under supraphysiological conditions. EGF receptor (EGFR) is up-regulated in a wide variety of tumors and hence is a target for cancer therapeutics. Abnormal expression/activation of EGFR contributes to keratinocytic carcinomas, and mice with reduced PS dosages have been shown to develop skin tumors. Here we demonstrate that the levels of PS and EGFR in the skin tumors of PS1(+/-)/ PS2(-/-) mice and the brains of PS1/2 conditional double knockout mice are inversely correlated. Deficiency in PS/gamma-secretase activity or APP expression results in a significant increase of EGFR in fibroblasts. Importantly, we show that AICD mediates transcriptional regulation of EGFR. Furthermore, we provide in vivo evidence demonstrating direct binding of endogenous AICD to the EGFR promoter. Our results indicate an important role of PS/gamma-secretase-generated APP metabolite AICD in gene transcription and in EGFR-mediated tumorigenesis.
- [Show abstract] [Hide abstract]
ABSTRACT: Previous studies show that transient axonal glycoprotein-1, a ligand of amyloid precursor protein, increases the secretion of amyloid precursor protein intracellular domain and is involved in apoptosis in Alzheimer's disease. In this study, we examined the effects of transient axonal glycoprotein-1 on U251 glioma cells. 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay showed that transient axonal glycoprotein-1 did not inhibit the proliferation of U251 cells, but promoted cell viability. The terminal deoxynucleotidyl transferase dUTP nick end labeling assay showed that transient axonal glycoprotein-1 did not induce U251 cell apoptosis. Real-time PCR revealed that transient axonal glycoprotein-1 substantially upregulated levels of amyloid precursor protein intracellular C-terminal domain, and p53 and epidermal growth factor receptor mRNA expression. Thus, transient axonal glycoprotein-1 increased apoptosis-related gene expression in U251 cells without inducing apoptosis. Instead, transient axonal glycoprotein-1 promoted the proliferation of these glioma cells.Neural Regeneration Research 03/2014; 9(5):519-25. · 0.23 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: One of the most critical pathological features of Alzheimer's disease (AD) is the accumulation of β-amyloid (Aβ) peptides that form extracellular senile plaques in the brain. Aβ is derived from β-amyloid precursor protein (APP) through sequential cleavage by β- and γ-secretases. γ-secretase is a high molecular weight complex minimally composed of four components: presenilins (PS), nicastrin, anterior pharynx defective 1 (APH-1), and presenilin enhancer 2 (PEN-2). In addition to APP, γ-secretase also cleaves many other type I transmembrane (TM) protein substrates. As a crucial enzyme for Aβ production, γ-secretase is an appealing therapeutic target for AD. Here, we summarize current knowledge on the structure and function of γ-secretase, as well as recent progress in developing γ-secretase targeting drugs for AD treatment.Frontiers in Cellular Neuroscience 01/2014; 8:427. · 4.18 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Proteolytic processing of the amyloid-β protein precursor (AβPP) occurs via alternative pathways, culminating with the production of the AβPP intracellular domain (AICD). AICD can translocate to the nucleus and regulate transcription, but its activity is modulated by interactions with other proteins. In the nucleus, AICD, FE65, and Tip60 associate into AFT complexes, which are targeted to nuclear spots which correspond to transcription factories. Here we report that RanBP9 interacts with the cytoplasmic domain of AβPP, through the NPXY internalization motif. Moreover, RanBP9 interaction with Tip60 is also described. The RanBP9-Tip60 interaction dramatically relocated RanBP9 from a widespread cellular distribution to nuclear speckles. AβPP processing is a central aspect in determining the protein's function and that of its resulting proteolytic fragments, among them AICD. The latter results from the amyloidogenic pathway and is the peptidic species predominantly involved in nuclear signaling. Of note RanBP9 transfection was previously demonstrated to increase amyloid-β generation. Here we show that RanBP9 relocates AICD to the Tip60-enriched nuclear speckles, and prevented the formation of nuclear spots formation, having therefore a negative effect on AICD mediated nuclear signaling and consequently AFT complex formation. Furthermore, by transfecting cells with increasing amounts of RanBP9, the expression of AICD-regulated genes, including AβPP itself, was reduced. Given the data presented, one can deduce that RanBP9 has an inhibitory regulatory effect on AICD-mediated transcription and the effect is mediated by relocating AICD away from transcription factories.Journal of Alzheimer's disease: JAD 07/2014; · 3.61 Impact Factor
of ?-amyloid precursor protein regulates
EGF receptor expression
Yun-wu Zhang*†‡, Ruishan Wang*†, Qiang Liu§, Han Zhang*†, Francesca-Fang Liao*, and Huaxi Xu*†¶
*Center for Neuroscience and Aging, Burnham Institute for Medical Research, La Jolla, CA 92037;†Institute for Biomedical Research, Xiamen
University, Xiamen 361005, China; and§Department of Pediatrics, Washington University School of Medicine, St. Louis, MO 63110
Communicated by S. J. Singer, University of California at San Diego, La Jolla, CA, April 27, 2007 (received for review March 5, 2007)
Presenilins (PS, PS1/PS2) are necessary for the proteolytic activity
of ?-secretase, which cleaves multiple type I transmembrane pro-
teins including Alzheimer’s ?-amyloid precursor protein (APP),
Notch, ErbB4, etc. Cleavage by PS/?-secretase releases the intra-
cellular domain (ICD) of its substrates. Notch ICD translocates into
the nucleus to regulate expression of genes important for devel-
opment. However, the patho/physiological role of other ICDs,
especially APP ICD (AICD), in regulating gene expression remains
controversial because evidence supporting this functionality stems
mainly from studies performed under supraphysiological condi-
tions. EGF receptor (EGFR) is up-regulated in a wide variety of
tumors and hence is a target for cancer therapeutics. Abnormal
expression/activation of EGFR contributes to keratinocytic carci-
nomas, and mice with reduced PS dosages have been shown to
develop skin tumors. Here we demonstrate that the levels of PS
and EGFR in the skin tumors of PS1?/?/PS2?/?mice and the brains
Deficiency in PS/?-secretase activity or APP expression results in a
significant increase of EGFR in fibroblasts. Importantly, we show
that AICD mediates transcriptional regulation of EGFR. Further-
more, we provide in vivo evidence demonstrating direct binding of
endogenous AICD to the EGFR promoter. Our results indicate an
gene transcription and in EGFR-mediated tumorigenesis.
Alzheimer’s disease ? ?-amyloid precursor protein intracellular domain ?
transcriptional regulation ? tumorigenesis
tease consisting of at least three other components: nicastrin
(Nct), APH-1 (anterior pharynx-defective-1), and PEN-2 (pre-
senilin enhancer-2) (1, 2). PS/?-secretase is widely considered as
a potential target for developing therapies against Alzheimer’s
disease, because it is critical for the generation of ?-amyloid that
is pivotal in Alzheimer’s disease pathogenesis (3). Besides
cleaving ?-amyloid precursor protein (APP), PS/?-secretase has
a wide spectrum of type I membrane protein substrates including
Notch, ErbB4, CD 44, nectin-1?, E-cadherin, and low-density
lipoprotein receptor-related protein (1, 2). Cleavage of Notch by
PS/?-secretase releases Notch intracellular domain (NICD),
which can translocate into the nucleus and regulate downstream
gene expressions that are important for development (4, 5).
However, the physiological roles of other intracellular domains
(ICDs) cleaved by PS/?-secretase from substrates other than
Notch have yet to be determined. Several recent studies have
suggested that APP ICD (AICD) has transactivation activity and
can regulate transcription of multiple genes including APP,
controversial (10), primarily because the key evidence support-
ing the functions of AICD in transcriptional regulation has
mainly come from in vitro experiments; a direct binding of AICD
to the promoter of any given gene under physiological conditions
has not yet been established.
resenilins, including two homologs, PS1 and PS2, function as
the catalytic subunit of ?-secretase, an intramembrane pro-
Abolishment of PS/?-secretase activity by targeted disruption
of PS (11, 12), nicastrin (13), or APH-1 (14) genes in mice results
in embryonic lethality with defects resembling those found in
Notch-null mouse embryos (15, 16), making it difficult to
evaluate the additional physiological functions of PS/?-secretase
and its substrate metabolites in the adults. To circumvent this
obstacle, alternative approaches have been applied, including
the construction of a conditional double knockout (DKO)
mouse model lacking both PS1 and PS2 expression in the
postnatal forebrain (17), the neuron-specific expression of hu-
man PS1 in PS1-null mice for rescue of lethality (18), and the
generation of mice with reduced PS gene dosage that are
heterozygous for PS1 and null for PS2 (PS1?/?/PS2?/?) (19).
These studies have revealed that PSs play important roles in
synaptic plasticity and neuronal survival, in tumorigenesis, and
in hematopoiesis (17, 19, 20). In the mouse model of the human
PS1-rescued PS1-null mice, skin tumor phenotypes emerge
during aging (20). It has been suggested that enhanced ?-catenin
may be involved in skin tumorigenesis in these mice (20). In
addition, Nct heterozygous mice with 50% reduction of ?-
secretase have also been found to have increased risk of devel-
oping squamous cell carcinoma resembling that of human head
and neck squamous carcinoma (P. Wong, personal communica-
tion). However, PS/?-secretase deficiency may also distort other
signaling pathways involved in tumorigenesis, which deserves
EGF receptor (EGFR) is a protein tyrosine kinase that
belongs to the ErbB/HER family. The level of EGFR has been
found elevated in multiple tumor types, and overexpression of
EGFR correlates with poor clinical prognosis and tumor resis-
tance to chemotherapy. Hence, inhibition of EGFR is an im-
portant issue for cancer therapeutics (21). However, anti-EGFR
therapy shows its most common side-effect as skin toxicity
because EGFR signaling is essential to normal keratinocyte
biology such as cell cycle, proliferation, differentiation, and
survival (22). Dysregulation of EGFR expression and activation
has also been found to be involved in hyperproliferative skin
diseases (23). Because mice deficient in PS/?-secretase activity
tend to develop skin tumors, we investigated potential correla-
Author contributions: Y.-w.Z., F.-F.L., and H.X. designed research; Y.-w.Z., R.W., Q.L., and
H.Z. performed research; Y.-w.Z. and H.X. analyzed data; and Y.-w.Z. and H.X. wrote the
The authors declare no conflict of interest.
Freely available online through the PNAS open access option.
Abbreviations: AD, Alzheimer’s disease; APP, ?-amyloid precursor protein; APLP, APP-like
protein; ICD, intracellular domain; AICD, APP ICD; NICD, Notch ICD; KO, knockout; DKO,
double KO; EGFR, EGF receptor; PS, presenilin; Nct, nicastrin.
‡To whom correspondence may be sent at the † address. E-mail: firstname.lastname@example.org.
¶To whom correspondence may be sent at the*address. E-mail: email@example.com.
This article contains supporting information online at www.pnas.org/cgi/content/full/
© 2007 by The National Academy of Sciences of the USA
www.pnas.org?cgi?doi?10.1073?pnas.0703903104 PNAS ?
June 19, 2007 ?
vol. 104 ?
no. 25 ?
found that PS/?-secretase deficiency indeed results in an eleva-
tion of EGFR level in both the skin and the brain. More
importantly, we demonstrate that PS/?-secretase regulates
directly binds to EGFR promoter and regulates EGFR gene
Results and Discussion
PS/?-Secretase Deficiency Results in an Elevation of EGFR Level. Mice
with reduced PS gene dosage that are heterozygous for PS1 and
null for PS2 (PS1?/?/PS2?/?) develop splenomegaly with severe
granulocyte infiltration (19). Interestingly, these mice also de-
velop skin tumors during aging (Fig. 1A), a phenomenon very
similar to that found in the human PS1-rescued PS1-null mice
(20). We investigated the levels of EGFR in individual skin
tumor samples from these mice. Our results showed that the
steady-state level of EGFR inversely correlates with PS1 expres-
sion in the individual tumors (Fig. 1B). To confirm that PSs
indeed mediate the level of EGFR, we further analyzed brain
samples of mice lacking both PS1 and PS2 expression in the
postnatal forebrain (PS cDKO). These mice exhibit impairments
in synaptic plasticity, memory and learning, and neuronal sur-
vival, suggesting essential roles of PSs in normal neuronal
functions (17). Similarly, we found that the levels of EGFR were
dramatically increased in the brains of PS cDKO mice (Fig. 1C).
in the embryonic fibroblast cells derived from PS1/PS2 DKO (PS
DKO) mice when compared with those derived from their
littermate controls (PS WT). As shown in Fig. 2A Left, the level
of EGFR in PS DKO cells was significantly higher than that in
PS WT cells. Moreover, the elevation of EGFR in PS DKO cells
can be reversed by the stable expression of exogenous human
PS1 (Fig. 2A Right), excluding the possibility of cell clonal
the cellular level of EGFR. To investigate whether PS/?-
secretase activity is required for regulating EGFR level, we
treated PS DKO cells stably expressing pcDNA (PS DKO/
pcDNA, as control) or human PS1 (PS DKO/hPS1) with
L-685,458, a potent PS/?-secretase inhibitor. Treatment with
L-685,458 had no effect on the levels of EGFR in PS DKO/
pcDNA cells in which ?-secretase activity is absent. On the other
hand, in PS DKO/hPS1 cells, L-685,458 treatment, which is
known to increase APP ? C-terminal fragments resulting from
impaired ?-secretase activity, was able to cause an elevation of
EGFR (Fig. 2B). In addition, Nct knockout (KO) cells deficient
of ?-secretase activity (13, 24) exhibited a markedly increased
level of EGFR, which can be reversed by exogenously overex-
pressing human Nct (Fig. 2C). These results suggest that the
observed up-regulation of EGFR is mediated by the deficiency
of PS/?-secretase activity.
In addition to its essential role in ?-secretase activity, PS1 has
been shown to regulate intracellular trafficking of a series of
proteins such as APP, Nct, TrkB, and telecephalin (1). Because
activation of EGFR on the cell surface is triggered by binding to
its ligand EGF, which in turn causes endocytosis of the receptor
(A) PS1?/?/PS2?/?mice develop skin tumors during aging. (Left Upper) An
18-month-old PS1?/?/PS2?/?mouse showing massive skin lesions compared
with its littermate control. (Left Lower) A representative skin lesion on the
back of an 18-month-old PS1?/?/PS2?/?mouse. (Right) Hematoxylin/eosin
staining of PS1?/?/PS2?/?mouse skin tumor showing expanded dermis with
infiltrating clusters of neoplastic squamous epithelial cells (i), which are
characteristic of locally invasive squamous cell carcinoma. (Scale bar: 250 ?m.)
(B) The level of EGFR is inversely correlated to the level of PS1 in tumors from
PS1?/?/PS2?/?mice. Equal amounts of protein lysates of six tumor samples
derived from different PS1?/?/PS2?/?mice were analyzed by electrophoresis
on 4–20% SDS/PAGE gels and immunoblotted with antibodies against EGFR,
PS1 N-terminal fragment, and ?-actin (as loading control). (C) The level of
PS cDKO mice at 6 months of age and three brain samples from littermate
controls were analyzed by SDS/PAGE and immunoblotted with antibodies
against EGFR and ?-tubulin. Quantification was done by comparing the
densitometric values. Data represent means ? SD of EGFR level normalized to
that of ?-tubulin and relative to that of control.*, P ? 0.041, PS cDKO vs.
control (n ? 3).
PS deficiency results in tumorigenesis and an increase in EGFR level.
level of EGFR is significantly increased in PS1/PS2 DKO (PS DKO) mouse
fibroblast cells, and the elevation of EGFR can be reversed by overexpressing
exogenous human PS1. Equal amounts of lysates from PS DKO and WT
human PS1 and control blank pcDNA vector were analyzed and immunoblot-
Inhibition of PS/?-secretase activity increases the level of EGFR. PS DKO cells
stably expressing pcDNA blank vector or PS1 were treated with 500 nM
L-685,458 (458) for 24–48 h. Equal amounts of lysates were analyzed by
SDS/PAGE and immunoblotted with antibodies against EGFR, APP C-terminal
the elevation of EGFR can be reversed by overexpressing exogenous human
Nct. Equal amounts of lysates from Nct WT cells and Nct KO cells and equal
amounts of lysates from Nct KO cells stably expressing pcDNA (control) and
human Nct were analyzed and immunoblotted with antibodies against EGFR,
Nct, and ?-tubulin. (D) The level of cell surface EGFR is increased in PS DKO
cells. PS WT and PS DKO cells were incubated in the presence (?) or absence
cell surface proteins were affinity-precipitated (AP) with streptavidin, ana-
lyzed, and immunoblotted with anti-EGFR antibody. Five percent of cell
lysates was loaded as controls. Data represent means ? SD from three inde-
PS-mediated ?-secretase activity regulates the level of EGFR. (A) The
www.pnas.org?cgi?doi?10.1073?pnas.0703903104Zhang et al.
(25), we analyzed the level of EGFR in the plasma membrane in
PS-deficient cells. Our results showed that the level of cell
surface biotinylated EGFR was also elevated accompanied by
proportion of cell surface EGFR to the total EGFR in PS WT
cells was greater than that found in PS DKO cells (Fig. 2D, lane
1 vs. lane 5 compared with lane 3 vs. lane 7), implicating reduced
EGFR trafficking to the cell surface in the absence of PS, which
is consistent with the established trafficking role of PS (1). The
endocytic rates of EGFR upon EGF activation in PS DKO cells
and the DKO cells expressing exogenous human PS1 were also
determined and found to be unchanged [supporting information
(SI) Fig. 6 and SI Methods], suggesting that PS deficiency does
not affect endocytosis of EGFR.
PS/?-Secretase Deficiency Increases Biosynthesis of EGFR. The ob-
served increase in the steady-state level of EGFR could be due
to increased biogenesis, decreased degradation, or both. To
address this issue, we performed pulse–chase experiments and
found that after 5, 10, and 15 min of pulse-labeling, PS DKO/
pcDNA cells synthesized significantly higher levels of EGFR, as
compared with PS DKO/hPS1 cells (Fig. 3A Top). The catabo-
lism patterns of EGFR in the absence and presence of PS1 were
similar during various periods of chase after 15 min of pulse-
labeling, suggesting that PS/?-secretase affects the biosynthesis
rather than the degradation of the receptor (Fig. 3A Middle and
Bottom). To investigate whether PS/?-secretase affects EGFR
biogenesis at the transcription level, we performed quantitative
real-time PCR analysis and found a marked increase of EGFR
mRNA in PS DKO cells (?9-fold) compared with that in PS WT
cells (Fig. 3B).
APP/AICD Regulates EGFR Level. APP is an important PS/?-
secretase substrate, and its proteolytic product ?-amyloid plays
a central role in Alzheimer’s disease pathogenesis (3). The APP
gene family also includes two additional members, APP-like
proteins 1 and 2 (APLP1 and APLP2), but their physiological
functions remain elusive. Mice deficient in the single APP family
member exhibit negligible phenotype, whereas APP/APLP2
DKO mice (APP/APLP2 DKO) are early postnatal lethal,
suggesting functional redundancy between APP family members
(26). All of the APP family members share a high degree of
sequence conservation at their ICDs, implying important func-
tions of this domain. Several recent studies have suggested that
AICD has transactivation activity and can regulate transcription
of multiple genes including APP, GSK-3?, KAI1, neprilysin, and
BACE (6–9). To explore the potential role of APP/AICD in the
up-regulation of EGFR by PS/?-secretase deficiency, we studied
the cellular level of EGFR in APP/APLP2 DKO fibroblast cells
and found a drastic increase of EGFR in APP/APLP2 DKO
cells compared with a low level of EGFR in APP WT control
cells. The elevation of EGFR can be reversed by stably express-
4A), indicating APP’s involvement in regulating EGFR.
We next tested which APP domain(s) or metabolite(s) was
AICD suppressed the elevation of EGFR in APP/APLP2 DKO
cells (Fig. 4B), whereas overexpression of soluble APP? or
treatment with ?-amyloid failed to reduce EGFR level in the
absence of APP/APLP2 (SI Fig. 7 and SI Methods). Multidomain
protein Fe65 has been shown to interact with and stabilize
AICD. Fe65/AICD complex has been postulated to translocate
into the nucleus and bind to histone acetyltransferase Tip60; the
Fe65/AICD/Tip60 ternary complex may then exert the transcrip-
expressing pcDNA or human PS1 were pulse-labeled with [35S]methionine for
5, 10, or 15 min (Top). In some experiments, cells were first pulse-labeled for
15 min and then chased for various time periods (Middle). Cell lysates were
immunoprecipitated with EGFR antibody, followed by SDS/PAGE analysis and
autoradiography to detect labeled EGFR. The EGFR band intensity in each of
autoradiograph from three separate experiments was quantitated (Bottom).
Data represent means ? SD. (B) PS deficiency increases the level of EGFR
mRNA. Total RNA was extracted from PS WT and PS DKO cells and reverse-
transcribed for RT-PCR. The level of EGFR mRNA was normalized to that of
?-actin and compared with that of controls (PS WT, defined as one arbitrary
unit).*, P ? 0.020, PS DKO vs. PS WT (n ? 3).
PS/?-secretase deficiency increases biosynthesis of EGFR. (A) PS defi-
DKO (APP/APLP2 DKO) cells. Equal amounts of lysates from APP WT cells, APP/
APLP2 DKO cells, and APP/APLP2 DKO cells stably expressing pcDNA (control) or
human APP were analyzed and immunoblotted with antibodies against EGFR,
by exogenously overexpressing AICD. APP/APLP2 DKO cells were transiently
transfected with Notch N?E, pcDNA (control), AICD, AICD plus Fe65, or Fe65
alone. Equal amounts of cell lysates were analyzed and immunoblotted with
antibodies against EGFR, myc (for AICD and N?E/NICD recognition), FLAG (for
Fe65 recognition), and ?-tubulin. Quantitation was done by comparing the
densitometric values. Data represent means ? SD of EGFR level normalized to
deficiency increases the level of EGFR mRNA. Total RNA was extracted from APP
WT and APP/APLP2 DKO cells and reverse-transcribed for RT-PCR. The relative
as one arbitrary unit).*, P ? 0.036 (n ? 3).
Zhang et al. PNAS ?
June 19, 2007 ?
vol. 104 ?
no. 25 ?
tional regulation functions (6, 27, 28). Consistent with these
reports, coexpression of Fe65 and AICD dramatically increased
the level of AICD and suppressed the elevation of EGFR to a
greater extent than did AICD overexpression alone (Fig. 4B).
Overexpressing Fe65 alone in APP/APLP2 DKO cells had
minimal effect on the level of EGFR (Fig. 4B). Because NICD
has been well documented to act as a transcription factor for
downstream gene expression regulation (4, 29), we also exam-
ined the possible effect of NICD by overexpressing mouse
Notch1 N?E cDNA into APP/APLP2 DKO cells. N?E lacks the
ectodomain of Notch1, which allows it to be processed in a
ligand-independent manner by PS/?-secretase to produce NICD
(Fig. 4B and ref. 5). In contrast to AICD, overexpression of
Notch N?E/NICD had no effect on the cellular level of EGFR
in cells lacking APP/AICD (Fig. 4B), suggesting that Notch/
NICD is not involved in the regulation of EGFR expression.
However, our conclusion does not contradict reported roles of
Notch/NICD in tumorigenesis, a multifaceted process that in-
volves multiple signaling events or pathways. Besides AICD and
(1, 2). It is highly possible that some of these ICDs may also
participate in regulating EGFR expression. This possibility is
mRNA level of EGFR in PS/?-secretase-deficient cells is higher
than that in AICD-deficient cells (9-fold vs. 5-fold; Figs. 3B and
4C, which were performed side by side) and requires further
AICD Binds to EGFR Promoter and Regulates EGFR Gene Expression.
Having shown that expression of AICD inversely correlates with
the cellular level of EGFR, we next studied whether AICD
directly regulates EGFR gene expression. We PCR-amplified a
1.2-kb fragment at the 5?-flanking region of the EGFR gene (SI
Fig. 8A) and subcloned it into a promoterless firefly luciferase
reporter plasmid. This fragment possesses promoter activity as
monitored by luciferase expression (SI Fig. 8B). When this
EGFR promoter fragment was cotransfected with AICD alone
or AICD plus Fe65 into APP/APLP2 DKO cells, its ability to
drive luciferase expression was dramatically reduced, strongly
suggesting that AICD negatively regulates EGFR promoter
activity (Fig. 5A). Fe65 alone or Notch/NICD had little effect on
EGFR promoter activity.
The key evidence supporting the functions of AICD in tran-
scriptional regulation has mainly come from in vitro experiments
studying promoter transactivation and is deemed controversial.
A direct binding of AICD to the promoter of any given gene
under the physiological condition has not been reported. Re-
cently, several studies using ChIP assay have shown that, when
APP is overexpressed together with Fe65 or Fe65/Tip60 in cells,
the Fe65/AICD/Tip60 ternary complex can bind to the KAI1
promoter (6, 30). To identify direct binding between the Fe65/
AICD/Tip60 complex and the EGFR promoter region in vivo
under the physiological condition, we performed ChIP assay
using brain lysates from WT C57/BL mice, in which endogenous
brain AICD has been shown to be detectable (31), and from
APP/APLP2 DKO mice. After immunoprecipitation with anti-
bodies against APP C-terminal region, Fe65 or Tip60, the
collected DNAs were used as templates for PCR to amplify the
1 and SI Fig. 8A). One pair of primers (EGFR-334-start/EGFR-
671-stop) generated positive results from C57/BL mouse brain
samples, showing that the DNA fragments immunoprecipitated
by all three antibodies but not by normal rabbit IgG (as negative
control) contain the EGFR promoter region (Fig. 5B Left),
proving that endogenous AICD can directly bind to the EGFR
promoter in vivo. There were no positive PCR products gener-
ated from APP/APLP2 DKO mouse brain samples when APP
antibody was used to immunoprecipitate the DNA (Fig. 5B
Right). In addition, we performed a ChIP assay using WT
embryonic fibroblast cells. As shown in Fig. 5C, both APP and
Fe65 antibody immunoprecipitated EGFR promoter, whereas
IgG failed to do so. Transcription factor AP-2?, which has been
known to bind to EGFR promoter (32), was used as the positive
control for ChIP assay.
In summary, we demonstrate an inverse correlation between
the level of EGFR and PS1/?-secretase activity involving the
transcriptional regulation of EGFR gene expression by the
intracellular APP proteolytic product AICD. Our findings put
forward the concept that ?-secretase may function as a tumor
suppressor through altering the EGFR pathway/signaling, which
underscores the limitations of targeting ?-secretase for diseases.
The potential function of AICD as a transcription factor has
previously been proposed but remains controversial. Our study
presented herein provides direct evidence that PS1/?-secretase-
generated AICD can bind to the EGFR promoter and negatively
regulate transcription of EGFR gene. The identification of a
mechanism by which biogenesis/metabolism of EGFR, a key
target for cancer therapy, can be negatively regulated by AICD
and PS/?-secretase activity may enrich our understanding of the
functions of APP/AICD, PS1/?-secretase actions, and EGFR-
Materials and Methods
Mice. PS1?/?/PS2?/?mice (19) and APP/APLP2 DKO mice (33)
were kindly provided by H. Zheng (Baylor College of Medicine,
Houston, TX). Mouse handling procedures were performed in
plasmid containing the EGFR promoter region was cotransfected with pcDNA
cells. The Renilla luciferase plasmid phRL-SV40 was also cotransfected for
normalization purposes. After 24–48 h, cells were lysed and the luciferase
activities were measured with a luminometer. Data represent means ? SD.*,
P ? 0.05 (n ? 4). (B) AICD directly binds to the EGFR promoter in vivo.
mouse brain tissues were sonicated and immunoprecipitated with normal
rabbit IgG or antibodies against AICD, Fe65, and Tip60. Immunoprecipitated
DNA was purified and used as template for PCR amplification of EGFR pro-
moter regions. Five percent of unimmunoprecipitated DNA was used as input
for PCR amplification. PCR products were resolved on 2% agarose gels and
promoter in cultured cells. Cross-linked chromatin extracts from WT embry-
onic fibroblast cells were sonicated and immunoprecipitated with normal
rabbit IgG or antibodies against AICD, Fe65, and AP-2?. Immunoprecipitated
DNA was used as template for PCR as described for B.
AICD binds to EGFR promoter and regulates its gene expression. (A)
www.pnas.org?cgi?doi?10.1073?pnas.0703903104Zhang et al.
accordance with Burnham Institute for Medical Research An-
imal Research Committee and National Institutes of Health
Tissues and Cell Cultures. Total brain lysate samples from PS1/PS2
conditional DKO (PS cDKO) mice and littermate controls at 6
School, Boston, MA). Embryonic fibroblast cells derived from
PS1/PS2 DKO (PS DKO) (a gift from B. De Strooper, Flanders
Interuniversity Institute for Biotechnology, Leuven, Belgium),
APP/APLP2 DKO (APP/APLP2 DKO) (a gift from H. Zheng
of Baylor College of Medicine), and Nct KO (a gift from P.
Wong, Johns Hopkins Medical Institute, Baltimore, MD) mice
as well as their respective controls were grown in high-glucose
DMEM supplemented with 10% FBS and antibiotics.
Histology. Skin tumor tissues from PS1?/?/PS2?/?mice were fixed
in 10% neutral buffered formalin for 24–48 h, dehydrated, and
stored in 70% ethanol at 4°C. Tissues were vacuum-embedded in
paraffin, sectioned at 5 ?m, and stained with hematoxylin/eosin.
Plasmids, Transfection, and Immunoblot.HumanPS1,APP695,and
Nct plasmids and mouse Notch1 N?E plasmid have previously
been described (5, 34–36). A cDNA fragment encoding the last
at the beginning was generated by PCR and inserted into
pcDNA3.1/myc-His (Invitrogen, Carlsbad, CA) between the
EcoRV and XbaI sites. The pcDNA-FLAG-Fe65 plasmid was
kindly provided by T. Suzuki (Hokkaido University, Sapporo,
Japan). Transfection was performed by using FuGENE 6
(Roche, Indianapolis, IN) or Lipofectamine 2000 (Invitrogen),
following the manufacturers’ instructions. For stable cell line
establishment, plasmids were cotransfected with the pAG3zeo
plasmid into cells and selected with zeocin. For Western blot,
immunoblotted with specific antibodies. Rabbit anti-PS1 N
terminus antibody Ab14, anti-APP C terminus antibody 369, and
anti-Nct antibody 716 were developed in our laboratory (24, 36,
37). Rabbit anti-EGFR antibody and mouse anti-myc antibody
were from Santa Cruz Biotechnology (Santa Cruz, CA). Mouse
anti-?-tubulin, mouse anti-?-actin, and rabbit anti-FLAG anti-
bodies were from Sigma (St. Louis, MO).
Cell Surface Protein Biotinylation. To biotinylate cell surface pro-
teins, cells were washed with ice-cold PBS containing 1 mM each
of CaCl2 and MgCl2 and incubated at 4°C with 0.5 mg/ml
Sulfo-NHS-LC-biotin (Pierce, Rockford, IL) for 20 min, and the
P-40 lysis buffer. After affinity precipitation with streptavidin
beads (Pierce), the biotinylated proteins were eluted, loaded
directly on gels for electrophoresis, and followed by Western blot
analysis with anti-EGFR antibody.
?-Secretase Inhibitor Treatment. ?-Secretase inhibitor L-685,458
treated with 500 nM L-685,458 for 24–48 h before analysis.
Pulse–Chase of EGFR. To assay EGFR metabolism, cells were
pulse-labeled with [35S]methionine (500 ?Ci/ml) for 5, 10, or 15
min at 37°C and collected for analysis. In some experiments, cells
were first labeled for 15 min and then washed with PBS and
with anti-EGFR antibody, followed by SDS/PAGE analysis and
Quantitative Real-Time PCR. Total RNA was extracted from cells
by using TRIzol reagent (Invitrogen). The SuperScript First-
Strand kit (Invitrogen) was used to synthesize first-strand cDNA
from the samples with an equal amount of RNA according to the
manufacturer’s instruction. Synthesized cDNAs were then am-
plified by using IQTM SYBR green supermix and ICycler from
Bio-Rad (Hercules, CA), and the data were analyzed by using
Bio-Rad MyIQ 2.0. Primers used for EGFR and ?-actin ampli-
fication were EGFR-RT-5?/EGFR-RT-3? and actin-5/actin-3,
respectively (see SI Table 1 for sequence of the primers). The
level of EGFR mRNA was normalized to that of ?-actin.
Luciferase Assay. We PCR-amplified a 5?-flanking region of the
mouse EGFR gene by using genomic DNA from PS DKO cells
as templates. Primers used were EGFR-21-start and EGFR-
1242-stop (SI Table 1). After amplification, PCR products were
ing. The EGFR fragment was then resubcloned into the pGL3-
enhancer vector containing the firefly luciferase gene (Promega,
Madison, WI). Firefly luciferase vectors were cotransfected with
phRL-SV40 containing the Renilla luciferase gene (Promega)
into cells for 24–48 h. Firefly luciferase activities were assayed
and normalized to those of Renilla luciferase.
ChIP. ChIP assays were performed by using a commercial kit
(Upstate, Chicago, IL) following the manufacturer’s instructions
with minor modifications. Briefly, the brain tissues from WT
C57/BL mice at 2–4 months of age or from perinatal APP/
with 1% formaldehyde in the tissue culture media to cross-link
proteins to DNA. Formaldehyde was also added directly into the
culture media for cross-linking in WT fibroblast cells. The cell
pellet was lysed and sonicated. After centrifugation, the super-
natant was incubated overnight at 4°C with antibodies against
APP C terminus (Invitrogen), Fe65 (Abcam, Cambridge, MA),
Tip60 (Upstate), or AP-2? (Cell Signaling, Danvers, MA), and
normal rabbit IgG (Upstate). After immunoprecipitation, the
antibody/protein/DNA complex was incubated at 65°C for 4 h to
reverse the protein/DNA cross-links. The DNA was purified and
used as a template for PCR amplification. Different pairs of
EGFR promoter primers were used for amplification (SI Table
1 and SI Fig. 8A). PCR products were resolved on 2% agarose
gels and visualized after ethidium bromide staining.
Statistical Analysis. Data were analyzed by using the two-tailed
Student t test for comparison of independent means.
We are grateful to Drs. H. Zheng (Baylor College of Medicine), J. Shen
(Harvard Medical School), B. De Strooper (Flanders Interuniversity
Institute for Biotechnology), P. Wong (Johns Hopkins Medical Insti-
tute), and T. Suzuki (Hokkaido University) for providing materials. This
work was supported in part by National Institutes of Health Grants R01
AG030197 (to H.X.), R01 NS046673 (to H.X.), R01 AG021173 (to
H.X.), and R01 NS054880 (to F.-F.L.); grants from the Alzheimer’s
Association and the American Health Assistance Foundation (to H.X.);
and a grant from the National Natural Science Foundation of China (No.
30672198 to Y.-w.Z.). Y.-w.Z. is the recipient of National Institutes of
Health Training Grant F32 AG024895.
1. Vetrivel KS, Zhang YW, Xu H, Thinakaran G (2006) Mol Neurodegener 1:4.
2. Iwatsubo T (2004) Curr Opin Neurobiol 14:379–383.
3. Greenfield JP, Gross RS, Gouras GK, Xu H (2000) Front Biosci 5:D72–D83.
4. Kopan R, Schroeter EH, Weintraub H, Nye JS (1996) Proc Natl Acad Sci USA
5. Schroeter EH, Kisslinger JA, Kopan R (1998) Nature 393:382–386.
6. Baek SH, Ohgi KA, Rose DW, Koo EH, Glass CK, Rosenfeld MG (2002) Cell
7. Kim HS, Kim EM, Lee JP, Park CH, Kim S, Seo JH, Chang KA, Yu E, Jeong
SJ, Chong YH, et al. (2003) FASEB J 17:1951–1953.
Zhang et al. PNAS ?
June 19, 2007 ?
vol. 104 ?
no. 25 ?
8. Pardossi-Piquard R, Petit A, Kawarai T, Sunyach C, Alves da Costa C, Vincent
B, Ring S, D’Adamio L, Shen J, Muller U, et al. (2005) Neuron 46:541–554.
9. von Rotz RC, Kohli BM, Bosset J, Meier M, Suzuki T, Nitsch RM, Konietzko
U (2004) J Cell Sci 117:4435–4448.
10. Hebert SS, Serneels L, Tolia A, Craessaerts K, Derks C, Filippov MA, Muller
U, De Strooper B (2006) EMBO Rep 7:739–745.
11. Shen J, Bronson RT, Chen DF, Xia W, Selkoe DJ, Tonegawa S (1997) Cell
12. Donoviel DB, Hadjantonakis AK, Ikeda M, Zheng H, Hyslop PS, Bernstein A
(1999) Genes Dev 13:2801–2810.
13. Li T, Ma G, Cai H, Price DL, Wong PC (2003) J Neurosci 23:3272–3277.
14. Ma G, Li T, Price DL, Wong PC (2005) J Neurosci 25:192–198.
15. Huppert SS, Le A, Schroeter EH, Mumm JS, Saxena MT, Milner LA, Kopan
R (2000) Nature 405:966–970.
16. Swiatek PJ, Lindsell CE, del Amo FF, Weinmaster G, Gridley T (1994) Genes
17. Saura CA, Choi SY, Beglopoulos V, Malkani S, Zhang D, Shankaranarayana
Rao BS, Chattarji S, Kelleher RJ, III, Kandel ER, Duff K, et al. (2004) Neuron
18. Qian S, Jiang P, Guan XM, Singh G, Trumbauer ME, Yu H, Chen HY, Van
de Ploeg LH, Zheng H (1998) Neuron 20:611–617.
19. Qyang Y, Chambers SM, Wang P, Xia X, Chen X, Goodell MA, Zheng H
(2004) Biochemistry 43:5352–5359.
20. Xia X, Qian S, Soriano S, Wu Y, Fletcher AM, Wang XJ, Koo EH, Wu X,
Zheng H (2001) Proc Natl Acad Sci USA 98:10863–10868.
21. Astsaturov I, Cohen RB, Harari P (2006) Expert Rev Anticancer Ther 6:1179–1193.
22. Sipples R (2006) Semin Oncol Nurs 22:28–34.
23. Jost M, Kari C, Rodeck U (2000) Eur J Dermatol 10:505–510.
24. Zhang YW, Luo WJ, Wang H, Lin P, Vetrivel KS, Liao F, Li F, Wong PC,
Farquhar MG, Thinakaran G, et al. (2005) J Biol Chem 280:17020–17026.
25. Citri A, Yarden Y (2006) Nat Rev Mol Cell Biol 7:505–516.
26. Zheng H, Koo EH (2006) Mol Neurodegener 1:5.
27. Cao X, Sudhof TC (2001) Science 293:115–120.
28. Kimberly WT, Zheng JB, Guenette SY, Selkoe DJ (2001) J Biol Chem
29. Kopan R, Goate A (2000) Genes Dev 14:2799–2806.
N, Rosenfeld MG, Russo T (2005) EMBO Rep 6:77–82.
31. Ryan KA, Pimplikar SW (2005) J Cell Biol 171:327–335.
32. Wang X, Bolotin D, ChuDH, Polak L, Williams T, Fuchs E (2006) J Cell Biol
33. Wang P, Yang G, Mosier DR, Chang P, Zaidi T, Gong YD, Zhao NM,
Dominguez B, Lee KF, Gan WB, et al. (2005) J Neurosci 25:1219–1225.
34. Lo AC, Haass C, Wagner SL, Teplow DB, Sisodia SS (1994) J Biol Chem
35. Vetrivel KS, Cheng H, Lin W, Sakurai T, Li T, Nukina N, Wong PC, Xu H,
Thinakaran G (2004) J Biol Chem 279:44945–44954.
36. Thinakaran G, Borchelt DR, Lee MK, Slunt HH, Spitzer L, Kim G, Ratovitsky
T, Davenport F, Nordstedt C, Seeger M, et al. (1996) Neuron 17:181–190.
37. Xu H, Sweeney D, Wang R, Thinakaran G, Lo AC, Sisodia SS, Greengard P,
Gandy S (1997) Proc Natl Acad Sci USA 94:3748–3752.
www.pnas.org?cgi?doi?10.1073?pnas.0703903104Zhang et al.