Báo cáo khoa học: Novel brain 14-3-3 interacting proteins involved in neurodegenerative disease doc

Novel brain 14-3-3 interacting proteins involved in
neurodegenerative disease
Shaun Mackie* and Alastair Aitken
University of Edinburgh, School of Biomedical and Clinical Laboratory Sciences, Edinburgh, Scotland, UK
Dimeric 14-3-3 proteins have important functions in
diverse biological processes [1–3]. An optimal motif for
14-3-3 binding was identified as R(S)XpSXP [4] where
Sp is phosphoserine. This was later refined to include a
second motif, mode 2, RXXXpSXP [5]. A number of
proteins also bind to 14-3-3 at their C-terminus where
the presence of a proline residue may be unnecessary
as the peptide backbone would not be required to loop
out again from the binding pocket. Thus in addition
to the well-characterized nonphosphorylated binding
motifs, there may be a third phospho-dependent 14-3-3-
binding motif, -pS ⁄ pT (X
1-2
)-CO
2
H, referred to by
Ganguly and colleagues as ‘mode III’ [6]. This motif
has also been characterized structurally in plant proton
ATPases [7]. The motif in b-COP (RRSpSV-CO
2
H)
may also come into this category [8]. Unphosphory-
lated motifs that interact with 14-3-3 at high affinity
have also been characterized [1,2].
Structures of 14-3-3 and the binding site of the
phospho- and unphosphorylated motifs have been
determined [2,5,6]. Phosphorylation of specific 14-3-3

isoforms can also regulate interactions [9].
14-3-3 isoforms are involved in neurodegenerative
disorders including Alzheimer’s [10] and Parkinson’s
disease [11]. We identified four of the seven mamma-
lian 14-3-3 isoforms (b, c, e and g) in the spinal fluid
(CSF) of patients with Creutzfeldt–Jakob disease
(CJD) [12]. 14–3-3 g alone was also present in all
patients with other dementias, including Alzheimer’s.
Changes in the localization of 14-3-3 isoforms were
Keywords
14-3-3; d-catenin; IRSp53;
neurodegenerative diseases; yeast two-
hybrid
Correspondence
A. Aitken, University of Edinburgh, School of
Biomedical and Clinical Laboratory Sciences,
George Square, Edinburgh, EH8 9XD,
Scotland, UK
Fax: +44 131 6503725
Tel: +44 131 6503721
E-mail:
*Present address
University of Edinburgh, Psychiatric Genet-
ics Section, Medical Genetics Section,
Western General Hospital, Edinburgh,
Scotland
(Received 31 January 2005, revised 13 May
2005, accepted 22 June 2005)
doi:10.1111/j.1742-4658.2005.04832.x
We isolated two novel 14-3-3 binding proteins using 14-3-3 f as bait in a

yeast two-hybrid screen of a human brain cDNA library. One of these
encoded the C-terminus of a neural specific armadillo-repeat protein,
d-catenin (neural plakophilin-related arm-repeat protein or neurojungin).
d-Catenin from brain lysates was retained on a 14-3-3 affinity column.
Mutation of serine 1072 in the human protein and serine 1094 in the equiv-
alent site in the mouse homologue (in a consensus binding motif for 14-3-
3) abolished 14-3-3 binding to d-catenin in vitro and in transfected cells.
d-catenin binds to presenilin-1, encoded by the gene most commonly
mutated in familial Alzheimer’s disease. The other clone was identified as
the insulin receptor tyrosine kinase substrate protein of 53 kDa (IRSp53).
Human IRSp53 interacts with the gene product implicated in dentatoru-
bral-pallidoluysian atrophy, an autosomal recessive disorder associated
with glutamine repeat expansion of atrophin-1.
Abbreviations
CJD, Creutzfeldt–Jakob disease; CSF, cerebrospinal fluid; DRPLA, dentatorubral-pallidoluysian atrophy; IRSp53, insulin receptor tyrosine
kinase substrate protein of 53 kDa; SCA1, spinocerebellar Ataxia Type 1.
4202 FEBS Journal 272 (2005) 4202–4210 ª 2005 FEBS
observed during neurodegeneration in mice as a result
of scrapie infection [13]. 14-3-3 isoforms play a key
role in neurodegeneration in the polyglutamine-repeat
disease spinocerebellar ataxia type 1 (SCA1) [14].
SCA1 is characterized by ataxia, progressive motor
deterioration and loss of cerebellar Purkinje cells
caused by the expansion of a region of the ataxin-1
gene that produces an abnormally long stretch of glu-
tamine. In SCA1, 14-3-3 f and e bind to and stabilize
ataxin-1, after phosphorylation by Akt, thus slowing
its normal degradation. A number of other inherited
neurodegenerative diseases, including Huntington’s
disease and Dentatorubral-pallidoluysian atrophy

(DRPLA) are caused by proteins that undergo a simi-
lar pathogenic polyGln expansion. 14-3-3 and a-synuc-
lein colocalize with the perinuclear inclusions of
huntingtin protein [15].
To identify novel 14-3-3 binding partners in mamma-
lian brain, we performed a two-hybrid screen with
human 14-3-3 f as bait and isolated clones for two pro-
teins involved in distinct neurodegenerative diseases.
Results
Identification of d-catenin as a 14-3-3 interacting
protein
The yeast two-hybrid screen of a human brain cDNA
library was carried out with a GAL4 binding domain
14-3-3 fusion protein as bait. From 2.54 · 10
6
trans-
formants screened, 35 diploid colonies (D1–D35) grew
up under selective conditions. 2 out of 35 colonies
specified in-frame coding region cDNAs.
BLAST searches showed that one of these, D16, was
homologous to the C-terminal region of delta catenin
(Primary accession number Q9UQB3 in Swiss-Prot,
also known as neural plakophilin related armadillo
protein, NPRAP or neurojungin) that is almost exclu-
sively expressed in the nervous system [16]. d-Catenin
is a member of the p120-catenin (p120ctn) subfamily,
defined as proteins with 10 armadillo (ARM) repeats
in characteristic spacing with diverse N- and C-ter-
minal flanking sequences [17,18] see Fig. 1A. The
42-residue repeated Arm motif was originally described

in the Drosophila segment polarity gene, armadillo [19].
The ARM domains of d-catenin are necessary and
sufficient for adherens junction targeting and for direct
interaction with cadherin (Fig. 1B).
Clone D16 encoded a putative protein product of
386 amino acids (839–1125) which included 4 of the
ARM repeats, a potential 14-3-3 binding site and a
PDZ binding motif (Fig. 1A).
Both northern blot and in situ hybridization studies
indicate that delta-catenin is almost exclusively
expressed in the nervous system [16,20]. d-Catenin has
a structure similar to that of p0071 and is considered
to be a neural isoform of p0071, which is expressed
ubiquitously [21].
β
or
γ−
catenin
p120ctn or
δ
-catenin
α
-catenin
β
or
γ−
catenin
p120ctn or
δ
-catenin

Intracellular
Extracellular
α
-catenin
Actin Filaments
Cadherin Receptor
Ca
++
PDZ binding motif
532
840
1225
RSApSAP
N
A
B
1
C
1225
1013
D16
2265bp
3’ UTR
14-3-3 phospho-binding siteARM domain
Fig. 1. d-Catenin domains. (A) Alignment of
d-catenin cDNA and protein domains with
clone D16. PCR amplification of the pACT2
D16 clone identified a  2200 nucleotide
insert. Sequence analysis established that
clone D16 encoded the C-terminus of

human d-catenin (GenBank accession num-
ber U96136) and  1kbof3¢-untranslated
region. The alignment of this fragment is
shown below the full-length human
d-catenin. The armadillo, ARM, domains,
predicted 14-3-3 binding site and a PDZ
binding motif are indicated. (B) d-catenin
complexes in adherens junction targeting
and interaction with cadherin.
S. Mackie and A. Aitken 14-3-3 in neurodegenerative disease
FEBS Journal 272 (2005) 4202–4210 ª 2005 FEBS 4203
14-3-3 Binds endogenous d-catenin from brain
lysates
To determine whether endogenous d-catenin associated
in brain tissue, we passed sheep brain homogenate
over GST-14-3-3 and control GST affinity columns.
d-Catenin specifically bound to a GST-14-3-3 f column
but not to a control GST column (Fig. 2A). We typic-
ally detected a doublet by western blot that may be
due to in vivo phosphorylation. A doublet has been
observed previously [22] and a splice variant of d-cate-
nin is known, although both forms include the 14-3-3
motif, which is not present in other, more widely
expressed catenins, suggesting that this interaction may
be functionally restricted to neuro-epithelial cells. We
also performed immunoprecipitation assays from sheep
and mouse brain homogenates using recombinant
GST-14-3-3 and detected a 160 kDa doublet band with
anti-d-catenin sera, consistent with the expected M
r

of
full-length d-catenin. The immunoprecipitations were
also probed with a phospho-specific antibody (New
England Biolabs) against the consensus RSXpSXP
14-3-3 binding motif (Fig. 2B). This specifically detec-
ted a 160 kDa polypeptide at the same position as the
d-catenin antibody suggesting that phosphorylation at
this site may be functionally important in vivo. Other
species were evident which may represent partially
degraded, phosphorylated forms of d-catenin.
A 14-3-3 binding site on d-catenin
Human and mouse d-catenin cDNAs encode proteins
of 1225 and 1247 residues, respectively, and share 95%
identity at the amino acid level [22]. A predicted 14-3-3
binding motif (RSApSAP) comprises phosphoSer1072
and 1094 and neighbouring residues, respectively.
Therefore to establish the mode of binding of 14-3-3
to d-catenin, the 386 residue proteins encoded by the
wild-type d-catenin clone D16 and the S1072A mutant
were expressed as
35
S-labelled proteins in IVTT. 14-3-3
interacted efficiently with wild-type d-catenin but not
with the S fi A mutant (Fig. 3). This indicates that
the interaction is phosphorylation dependent at this
site. We also used a synthetic phosphopeptide that
AB
Fig. 2. 14-3-3 Binds endogenous d-catenin from brain lysates.
Sheep brain homogenate was lysed in NaCl ⁄ P
i

buffer (including
protease inhibitors) containing 1% Triton-100 or NaCl ⁄ P
i
⁄ 1% TX100
plus 0.1% SDS to aid solubilization of brain d-catenin. The extract
was clarified by centrifugation at 40 000 g and the supernatant
passed through a GST-affinity column. The flow-through was
applied to a GST 14-3-3 f column. After extensive washing, bound
proteins were eluted by directly boiling of the glutathione–Seph-
arose beads in SDS ⁄ PAGE sample buffer and analysed by 6%
SDS ⁄ PAGE and blotting with rabbit anti-(d catenin) Ig (Ab62, from
K.S. Kosik, Harvard Medical School, Boston, USA). (A) Lane 1, con-
trol GST column (TX100, no SDS); Lane 2, lysate prepared in
1%TX100) and affinity purified on the GST zeta column; Lane 3,
lysate prepared in 1% TX100 plus 0.1% SDS to aid solubilization
and affinity purified on the GST zeta column. (B) Samples prepared
as in lanes 1 and 3 above then probed with anti-phospho 14-3-3 BS
monoclonal (NEB, Cell Signaling Technology).
A
B
Fig. 3. Mutation of Ser1072 of d-catenin abolishes binding to 14-3-
3. (A) Wild-type d-catenin clone D16 (S1072) and the S1072A
mutant were expressed as
35
S-labelled proteins in IVTT. Lanes 1,
2: Input (2%) of the two constructs; 3 and 4: immunoprecipitation
of wild-type D16 by GST and by GST-14-3-3. 5 and 6: immunopre-
cipitation of the D16 S1072A by GST and by GST-14-3-3. (B) A simi-
lar experiment was carried out in the presence of phosphorylated
and unphosphorylated peptides. Lanes 1, immunoprecipitation of

wild-type D16 by GST-14-3-3; 2, immunoprecipitation of wild-type
D16 by GST; 3, immunoprecipitation of D16 S1072A mutant by
GST-14-3-3; 4, immunoprecipitation of wild-type D16 by GST-14-3-3
in the presence of Raf-phosphopeptide (300 l
M); 5, immunoprecipi-
tation of wild-type D16 by GST-14-3-3 in the presence of the same
concentration of unphosphorylated peptide.
14-3-3 in neurodegenerative disease S. Mackie and A. Aitken
4204 FEBS Journal 272 (2005) 4202–4210 ª 2005 FEBS
corresponds to c-Raf1 amino acids 252–264, a canon-
ical 14-3-3 binding motif previously shown to dissoci-
ate Raf ⁄ 14-3-3 complexes [4]. By contrast, as a
control, the unphosphorylated version of this peptide
did not interfere with binding.
14-3-3 binds to d-catenin expressed in MDCK
cells
As cDNA encoding full-length human d-catenin was
unavailable, we used a mouse cDNA clone for sub-
sequent studies [16]. Classical adherens junctions hold
epithelial cells together via cadherin-catenin protein
complex linkages and d-catenin interacts with adhesive
junction proteins both in transfected cells and mouse
brain [22], Fig. 1B. Cadherins are Ca
2+
-dependent
cell-cell adhesion receptors involved in a variety of bio-
logical processes including development, morpho-
genesis and tumour metastasis. Cadherins on adjacent
cells contact one another through their extracellular
domains. The intracellular domains anchor the junc-

tional complex or adherens junction to the actin cyto-
skeleton via the cytoplasmic catenins.
Therefore to characterize 14-3-3 ⁄ d–catenin inter-
actions in a defined culture system where adhesive
junctions are prominent, we used Madin–Darby canine
kidney (MDCK) epithelial cells. Lysates were prepared
from cells transfected with untagged (as a control) and
FLAG-tagged d-catenin. GST-14-3-3 f immunoprecipi-
tations immunoblotted with anti-FLAG Ig, detected
a 160 kDa doublet which bound specifically to GST-
14-3-3 f. Specific interaction was observed with wild-
type full-length d-catenin in both Cos7 and MDCK
cells and was ablated by mutation of serine 1094 to
alanine (Fig. 4A,B).
To establish the site of binding of 14-3-3 to d-cate-
nin in MDCK cells, we performed binding assays in
the presence of competitor peptides. We again used the
synthetic c-Raf1 phosphopeptide (and the unphosphory-
lated version of this peptide as control, not shown)
and a nonphosphorylated peptide inhibitor of 14-3-3
binding, R18 (FHCVPRDLSWLDLEANMCLP). R18
was originally isolated from a phage display library
with high affinity for the phosphoserine-binding pocket
of 14-3-3 and which disrupts binding of 14-3-3 to tar-
get proteins such as Raf, Ask1 [23] and EXO-S [24].
Both peptides efficiently prevented 14-3-3 ⁄ d–catenin
complex association in vitro in cell extracts (Fig. 5).
These results verified that the interaction between
d-catenin and 14-3-3 is mediated through the phospho-
binding pocket of 14-3-3.

The 14-3-3 binding motif is not present in members
of the p120ctn sequence family, which are more ubi-
quitously expressed, suggesting that this interaction
may be functionally restricted to neuro-epithelial cells.
Interaction of IRSp53 with 14-3-3
We also identified a full-length clone of a 53 kDa SH3
domain-containing adaptor protein originally identified
as a substrate of the insulin receptor kinase (IRSp53).
IRSp53 interacts with Rho GTPases to regulate the
organization of the actin cytoskeleton and is a compo-
A
B
Fig. 4. 14-3-3 interacts with full-length d-catenin in cells. Cos7 (A)
and MDCK cells (B) were transiently transfected with either empty
vector (A) or untagged pcDNA wild type delta catenin (B) and Flag
tagged delta catenin wild type and Flag tagged d-catenin with a Ser
to Ala substitution at residue 1094 (S1094A). Transfected cell
extracts were split and incubated with 20 lg GST and 20 lg GST-
14-3-3 for 120 min at 4 °C. Upper panels: lysate loading controls.
Lanes: 1, untagged d-catenin; 2, Flag S1094 d-catenin; 3, Flag
S1094A d-catenin. Lower panels: Immunoprecipitation of Flag
tagged d-catenin (western blot with a-Flag). Lanes 1,3,5, GST
immunoprecipitation; lanes 2,4,6, GST-14-3-3 f immunoprecipita-
tion; Lanes 1,2, Vector (in Cos cells) or no flag tag (MDCK cells);
3,4, Flag S1094 d-catenin; 5,6, Flag S1094A d-catenin.
S. Mackie and A. Aitken 14-3-3 in neurodegenerative disease
FEBS Journal 272 (2005) 4202–4210 ª 2005 FEBS 4205
nent of signaling pathways that control the formation
of lamellipodia and filopodia [25].
A number of splice variants of human and mouse

IRSp53 are known, comprising mainly of a 12 residue
longer C-terminus and a 40 residue insertion around
residue 300 [26]. In this study we isolated the longer
form that is mainly expressed in brain. A ‘Scansite’
search [27] revealed a number of potential suboptimal
14-3-3 binding motifs in IRSp53 conserved across
mammalian species (Fig. 6). The best motif was the
medium stringency site at Ser512, RSVS
512
SG, which
would explain loss of binding of construct 1–366 but
motif(s) near the N-terminus must also be important
(e.g. RYLS
117
AA and ⁄ or RKKS
148
QG). A nonphos-
phorylated Ser immediately following Arg within the
first mode and a Pro two residues C-terminal to the
phosphorylated Ser or Thr in both motifs is strongly
favoured, but not an essential requirement for binding
to 14-3-3 [3].
As it was not clear which region or potential
motif(s) in IRSp53 might be involved in 14-3-3 associ-
ation we attempted to identify the site(s) of interaction
by deletion analysis. The constructs depicted in Fig. 7
A were coexpressed with HA-14–3-3f in Cos7 cells.
The results in Fig. 7B clearly indicate that deletion of
either the C-terminal region or the N- terminus caused
loss of 14-3-3 interaction. This may be due to a

requirement for binding through two sites to a 14-3-3
dimer and this type of tandem 14-3-3 binding has been
clearly shown to be functionally important in cases
such as Raf kinase [3,28] and the Forkhead transcrip-
tion factor FOXO4 [29].
It is also probable that the interaction between
14-3-3 and IRSp53 is not phosphorylation dependent
as treatment with lambda phosphatase of Cos7 cell
lysates, into which Flag-IRSp53 and HA-14-3-3 zeta
had been co transfected, did not reduce interaction
(Fig. 8A).
Immunoprecipitation experiments with a construct
with mutations in essential residues of the phospho-
peptide binding pocket, HA-14-3-3 zeta (R56A,
R60A), that was transfected in Cos7 cells showed that
much less IRSp53 was immunoprecipitated. This veri-
fied that IRSp53 interacts with 14-3-3 in the binding
pocket.
Discussion
One of the 14-3-3 interacting clones that we identified
in the 2-hybrid analysis encoded the armadillo repeat
protein named delta-catenin, NPRAP (neural plako-
philin-related arm-repeat protein) or neurojungin [16].
Fig. 5. 14-3-3 Binding peptides prevent 14-3-3 ⁄ d–catenin associ-
ation. MDCK cells expressing d-catenin constructs were lysed and
extracts incubated in the absence or presence of 300 m
M Raf phos-
phopeptide or nonphosphorylated peptide, R18, at 4 °C for 60 min
20 lg GST-14-3-3 was added for 120 min. GST fusions were recov-
ered on GSH–Sepharose beads and washed four times with 0.5 mL

lysis buffer. Samples were separated by 6% SDS PAGE and assayed
for associated d-catenin by anti-Flag immunoblots. Lanes 1–4, lysate
loading controls. Lane 1, no Flag; 2–4, Flag d-catenin; 5, no Flag, no
peptide; 6, Flag d-catenin, no peptide; 7, Flag d-catenin, plus Raf
phosphopeptide; 8, Flag d-catenin, plus R18.
Fig. 6. Alignment of IRSp53 cDNA and domains in the yeast two-hybrid clone, D6. (A) Domain alignment of full-length human IRSp53 cDNA.
The SH3 domains (residues 377–435) which bind atrophin-1; an autoinhibitory region (AIR) that regulates Cdc42 binding to the CRIB motif
and a Cdc42 binding motif (residues 238–292) and potential 14-3-3 binding sites are indicated. (B) PCR amplification of the pACT2 D6 clone
identified a 2.4 kb insert that encoded the complete IRSp53 cDNA insert and  1kbof3¢-untranslated region.
14-3-3 in neurodegenerative disease S. Mackie and A. Aitken
4206 FEBS Journal 272 (2005) 4202–4210 ª 2005 FEBS
d-Catenin was originally identified by its ability to bind
to the loop region of presenilin-1, encoded by the gene
most commonly mutated in familial Alzheimer’s disease
[30]. Presenilin-1 interacts with complexes including
d-catenin to modulate Wnt signaling which is respon-
sible for a variety of signaling events that lead to neural
plate formation and patterning decisions ⁄ development
in the embryonic nervous system [20].
Although there is no evidence that 14-3-3 zeta bind-
ing plays a role in this pathway, it nevertheless suggests
another link between 14-3-3 and neurodegenerative
disease. Wnt signaling also regulates neuronal
cytoskeleton structure, cerebellar synaptic differentia-
tion, apoptosis and degenerative processes in the aging
brain. The latter establishes a link to pathogenesis
in Alzheimer’s disease. Mutations in presenilin 1
(PS1) gene are the most common cause of early onset
familial Alzheimer’s disease. d-Catenin expression is
decreased in presenilin-1 deficient mice [30].

The other novel 14-3-3 interacting protein in our
study was identified from a full-length clone of the
insulin receptor tyrosine kinase substrate protein of
53 kDa (IRSp53). IRSp53 is an SH3 domain-contain-
ing adaptor protein originally identified as a substrate
of the insulin receptor kinase [25]. IRSp53 interacts
with Rho GTPases to regulate the organization of
the actin cytoskeleton and is a component of signaling
pathways that control the formation of lamellipodia
and filopodia [31]. Human IRSp53 was isolated as
a protein which interacts with the gene product
implicated in DRPLA, an autosomal recessive disorder
caused by CAG ⁄ glutamine repeat expansion of atro-
phin-1 [32]. While the DRPLA gene is ubiquitously
expressed, neuron death occurs in specific anatomical
areas of the brain.
In a yeast two-hybrid screen of a human foetal brain
cDNA library with a fragment of atrophin-1 (residues
335–1185, containing 10 CAG repeats) clones isolated
included IRSp53, hDVL1, d-Catenin and 14-3-3 [33].
A proline rich region near the polyGln tract of atro-
phin-1 bound to the SH3 domain of IRSp53 in vitro.
Our results therefore expand the range of interacting
proteins and diversity of neurodegenerative disorders
A
B
Fig. 7. Domains of IRSp53 interacting with 14-3-3. (A) Schematic of
the constructs of flag tagged IRSp53. (B) The ability of the N- and
C-terminal constructs of flag tagged IRSp53 to be immunoprecipi-
tated by HA-tagged 14-3-3 f. The constructs of flag tagged IRSp53

depicted in A were cotransfected with HA-14-3-3 f in Cos7 cells
and immunoprecipitated with anti-HA-Ig as described in Experimen-
tal procedures. Lane 1, HA-14-3-3 f + IRSp53; 2, HA-14-3-3 f +
IRSp53-FLAG; 3, HA-14-3-3 f + D 1–125 IRSp53-FLAG; 4, HA-14-3-
3 f + D 1–179 IRSp53-FLAG; 5, HA-14-3-3 f + D 1–366 IRSp53-
FLAG. Upper panel, expression levels of the constructs. Middle
panel, expression levels of HA-14-3-3 f. Lower panel, western blot
of IP with anti-Flag Ig.
A
B
Fig. 8. IRSp53 interacts with 14-3-3 in the binding pocket but may
do so in a nonphospho-dependent manner. (A) Flag-IRSp53 and
HA-14-3-3 zeta were co transfected into Cos7 cells as described in
Fig. 4 and Experimental procedures. The cell lysates were treated
with lambda phosphatase and immunoprecipitated with anti-HA Ig.
The IPs were western blotted with anti-Flag Ig. Lane 1, input; 2,
no phosphatase treatment; 3, treatment with lambda phosphatase.
(B) Flag-IRSp53 and HA-14-3-3 zeta constructs were co transfected
into Cos7 cells as described in Fig. 4 and Experimental procedures.
The cell lysates were immunoprecipitated with anti-HA Ig. The
pellets were western blotted with anti-Flag Ig. Lane 1, input; 2,
immunoprecipitation with wild-type HA-14-3-3 zeta; 3, immunopre-
cipitation with HA-14-3-3 zeta (R56A, R60A) mutant construct.
S. Mackie and A. Aitken 14-3-3 in neurodegenerative disease
FEBS Journal 272 (2005) 4202–4210 ª 2005 FEBS 4207
in which isoforms of 14-3-3 are implicated, including
another polyglutamine expansion disease, DRPLA.
The key feature of all these diseases is the accumula-
tion in specific areas of the brain of abnormal forms
of proteins which results in neurodegeneration. The

proteins that accumulate (due to their misfolding
and ⁄ or genetic mutation) are specific to each disease.
However, a common feature that is now emerging is
the involvement of specific isoforms of 14-3-3. Deter-
mining the component proteins and role of 14-3-3
complexes, may lead to advances in understanding of
how these protein complexes regulate brain functions.
Experimental procedures
Two-Hybrid Screen cDNA encoding the 14-3-3 f ORF was
cloned into the NdeI ⁄ Bam HI sites of the vector pGBKT7
(Clontech, Basingstoke, UK) to create an in-frame fusion
with the DNA binding domain of GAL4. Plasmid
pGBKT7 ⁄ 14-3-3 f was transformed into yeast strain
SFY526 and combined with a pretransformed Matchmaker
cDNA library (Clontech) using standard yeast mating pro-
cedures. Diploid colonies were selected for activation of his-
tidine (His) and adenine (Ade) reporter genes by growth on
SD medium lacking Ade, His, Leu and Trp for 7–10 days.
Clones that survived repeated auxotrophic selection were
assayed for b-galactosidase activity by use of 5-bromo-
4-chloro-3-indolyl-b-d-galactopyranoside (X-gal) as a sub-
strate. Plasmid DNA was isolated from 39 positive clones
and library inserts were amplified by PCR using pACT2
vector specific primers.
Plasmids and constructs
Mouse d-catenin in pcDNA3.1 was from W. Franke (Ger-
man Cancer Research Center, Heidelberg, Germany).
Plasmids pGEX-2T 14-3-3 f and HA tagged 14-3-3 f
(pcDNA.1 Zeo) have been described previously [35] The
d-catenin ORF was amplified by PCR and the product inser-

ted into Not1 ⁄ Xho1 cut pCMV TAG4A (Invitrogen, Paisley,
UK) to generate a C-terminal FLAG tagged construct. Site-
specific mutations were introduced into the d-catenin ORF
using the QuikChange Site Directed Mutagenesis System
(Stratagene, Cleveland, OH, USA) according to the manu-
facturer’s instructions and confirmed by DNA sequencing.
PCR amplification of DNA fragments was carried out using
Pfu Turbo DNA polymerase (Stratagene) and integrity of
cloned inserts were confirmed by DNA sequencing.
The construct with mutations in essential residues of the
phosphopeptide binding pocket, HA-14-3-3 zeta (R56A,
R60A) was generated by Stratagene Quickchange site direc-
ted mutagenesis according to the manufacturers instructions.
In vitro transcription and translation (IVTT) was carried
out using TnT expression kits (Promega) as described [34].
Transfection of cultured cells
Cos7 and MDCK cells (ATCC) were maintained in high glu-
cose Dulbecco’s modified Eagle’s medium (Sigma) supple-
mented with 10% foetal bovine serum (Life Technologies),
1· nonessential amino acid supplement, 1· glutamine, peni-
cillin and streptomycin (Life Technologies) in air plus 5%
CO
2
with constant humidity. Cells were transfected at
80–90% confluence using Lipofectamine 2000 (Invitrogen)
according to the manufacturer’s instructions and harvested
24–30 h later.
Protein extracts
Transfected cells (100 mm plates) were washed once with
NaCl ⁄ Pi, scraped into 1.8 mL lysis buffer (50 mm Tris-Cl

pH 7.5, 150 mm NaCl, 1% TX-100, I mm EDTA, 1 mm
dithiothreitol, 1 mm NaVO
4
,10mm NaF and protease
inhibitor cocktail without EDTA (Roche Molecular Bio-
chemicals), incubated on ice for 15 min and clarified by
centrifugation for 20 mins at 16 000 g in a refrigerated
microfuge. For brain extract preparation, adult sheep brain
was briefly rinsed in lysis buffer (50 mm Tris ⁄ Cl pH 7.5,
150 mm NaCl, 1% TX-100, 2 mm EDTA, 10% glycerol,
2mm dithiotreitol, 2 mm NaVO
4
,50mm NaF, 20 mm
b-glycerophosphate, 1 mm PMSF and 2x protease inhibitor
cocktail without EDTA) and homogenized in the same
buffer. Lysates were precleared at 13 000 g for 30 min at
4 °C. Supernatant from the lysates was further clarified by
centrifugation at 40 000 g for 60 min at 4 °C. Supernatants
were filtered through 0.2 lm syringe filters (Nalgene) before
application to GST or GST 14-3-3 f affinity columns.
Treatment of cell lysates with lambda phosphatase (New
England Biolabs) was with 400 units phosphatase
for 60 min at 30 °C, according to the manufacturers
instructions.
Immunoprecipitation
Equal amounts of GST or GST 14-3-3 f fusion protein
(20 lg) were incubated overnight at 4 °C on a rotary wheel
with lysates prepared from 10 cm dishes of confluent Cos7
or MDCK cells. Complexes were captured by incubation
with glutathione Sepharose beads for 2 h at 4 °C. After

centrifugation, beads were washed four times with lysis buf-
fer. Bound proteins were eluted with SDS sample buffer
and subjected to SDS PAGE and immunoblotting. For
immunoprecipitations, 5 lg anti-HA7 monoclonal Ig (Sig-
ma) was incubated for 4 h or overnight at 4 °C with con-
trol or HA expressing lysates. Immunocomplexes were
incubated with protein A ⁄ G beads (Pierce) for 2 h at 4 °C
and captured by centrifugation. Immunocomplexes were
washed as above before immunoblot analysis. SDS PAGE
and western blotting were performed by standard methods.
14-3-3 in neurodegenerative disease S. Mackie and A. Aitken
4208 FEBS Journal 272 (2005) 4202–4210 ª 2005 FEBS
Anti-FLAG M2 peroxidase conjugate and Anti-HA7 Igs
were from Sigma and signals were detected using ECL,
chemiluminescence detection (Amersham Pharmacia Bio-
tech, Buckinghamshire, UK).
Recombinant protein purification
Protein expression was induced in E. coli strain BL21(DE3)
(Novagen, Merck Biosciences, Nottingham, UK) carrying
plasmid pGEX-2T or pGEX-2T 14-3-3 f. Briefly, cultures
were grown overnight at 37 °C in Liquid Broth medium (Life
Technologies, Inc., Paisley, UK) containing 50 lgÆmL
)1
ampicillin and diluted the following day (1 ⁄ 10) in the same
medium. Culture growth continued at 30 °C until the absorb-
ance (600 nm) reached 0.8 to 1.0. Expression of the tagged
protein was induced by the addition of 0.5 mm isopropyl b-
d-thiogalactopyranoside for 3 h at 25 °C. The fusion
proteins were purified by affinity chromatography on gluta-
thione-Sepharose beads (Amersham Pharmacia Biotech.).

For large-scale preparation of GST and GST 14-3-3 f affinity
columns, fusion protein lysates prepared from 2.5 L of
induced bacterial culture ( 7 mg fusion protein) were used
to saturate 2 mL columns of glutathione-Sepharose beads.
Peptide competition studies: dissociation
of 14-3-3 ⁄ d-catenin complexes in vitro
Cell extracts were incubated with 300 l m synthetic phos-
phopeptide corresponding to a c-Raf1 14-3-3 binding motif
(residues 252–264, SQRQRSTpSTPNVH) as well as with
the control peptide of the same sequence but unphosphory-
lated or with 300 lm of a nonphosphorylated peptide (R18,
FHCVPRDLSWLDLEANMCLP [23].
Acknowledgements
The work was funded by a MRC programme grant to
AA. We thank Bengt Hallberg for the R18 peptide.
Mouse d-catenin in pcDNA3.1 was a kind gift from
the laboratory of Dr W. Franke; Rabbit anti-(d cate-
nin) Ig (Ab62), raised against residues 434–530 was a
kind gift from the laboratory of Dr Kenneth Kosik.
HA tagged 14–3-3f (pcDNA.1 Zeo) was from Preeti
Kerai and Thierry Dubois.
References
1 Aitken A (2002) Functional specificity in 14-3-3 isoform
interactions through dimer formation and phosphoryla-
tion. Chromosome location of mammalian isoforms and
variants. Plant Mol Biol 50, 993–1010.
2 Fu H, Subramanian RR & Masters SC (2000) 14-3-3
proteins: structure, function and regulation. Annu Rev
Pharmacol Toxicol 40, 617–647.
3 Yaffe MB (2002) How do 14-3-3 proteins work? - Gate-

keeper phosphorylation and the molecular anvil hypo-
thesis. FEBS Lett 513 , 53–57.
4 Muslin AJ, Tanner JW, Allen PW & Shaw AS (1996)
Interaction of 14-3-3 with signalling proteins is mediated
by the recognition of phosphoserine. Cell 84, 889–897.
5 Yaffe MB, Rittinger K, Volinia S, Caron PR, Aitken A,
Leffers H, Gamblin SJ, Smerdon SJ & Cantley LC
(1997) The structural basis for 14-3-3: phosphopeptide
binding specificity. Cell 91 , 961–971.
6 Ganguly S, Weller JL, Ho A, Chemineau P, Malpaux B
& Klein DC (2005) Melatonin synthesis: 14–3-3-depend-
ent activation and inhibition of arylalkylamine N-acetyl-
transferase mediated by phosphoserine-205. Proc Natl
Acad Sci USA 102, 1222–1227.
7 Wurtele M, Jelich-Ottmann C, Wittinghofer A &
Oecking C (2003) Structural view of a fungal toxin
acting on a 14-3-3 regulatory complex. EMBO J 22,
987–994.
8 O’Kelly I, Butler MH, Zilberberg N & Goldstein SA
(2002) Forward transport. 14-3-3 binding overcomes
retention in endoplasmic reticulum by dibasic signals.
Cell 111, 577–588.
9 Aitken A, Howell S, Jones D, Madrazo J & Patel Y
(1995) 14-3-3 a and d are the phosphorylated forms of
Raf-activating 14-3-3 b and f. In vivo stoichiometric
phosphorylation in brain at a Ser-Pro-Glu-Lys motif.
J Biol Chem 270, 5706–5709.
10 Layfield R, Fergusson J, Aitken A, Lowe J, Landon M
& Mayer J (1996) Neurofibrillary tangles of Alzheimer’s
disease brains contain 14-3-3 proteins. Neurosci Lett

209, 57–60.
11 Perez RG, Waymire JC, Lin E, Liu JJ, Guo F & Zig-
mond MJ (2002) A role for alpha-synuclein in the regu-
lation of dopamine biosynthesis. J Neurosci 22, 3090–
3099.
12 Wiltfang J, Otto M, Baxter HC, Bodemer M, Steinacker
P, Bahn E, Kornhuber J, Kretzschmar HA, Poser S &
Aitken A (1999) Isoform pattern of 14-3-3 proteins in
the cerebrospinal fluid of patients with Creutzfeldt–
Jakob disease. J Neurochem 73, 2485–2490.
13 Baxter HC, Liu W-G, Aitken A & Fraser JR (2002)
Immunolocalisation of 14-3-3 isoforms in normal and
scrapie infected murine brain. J Neuroscience 109, 6–15.
14 Chen H-K, Fernandez-Funez P, Acevedo SF, Lam YC,
Kaytor MD, Fernandez MH, Aitken A, Skoulakis
EMC, Orr HT, Botas J & Zoghbi HY (2003) Interac-
tion of Akt-phosphorylated ataxin-1 with 14-3-3 medi-
ates neurodegeneration in spinocerebellar ataxia type 1.
Cell 113, 457–446.
15 Waelter S, Boeddrich A, Lurz R, Scherzinger E, Lueder
G, Lehrach H & Wanker EE (2001) Accumulation of
mutant huntingtin fragments in aggresome-like inclusion
bodies as a result of insufficient protein degradation.
Mol Biol Cell 12, 1393–1407.
S. Mackie and A. Aitken 14-3-3 in neurodegenerative disease
FEBS Journal 272 (2005) 4202–4210 ª 2005 FEBS 4209
16 Paffenholz R & Franke WW (1997) Identification and
localisation of a neurally expressed member of the
plakoglobin ⁄ armadillo multigene family. Differentiation
61, 293–304.

17 Peifer M, Berg S & Reynolds AB (1994) A repeating
amino acid motif shared by proteins with diverse cellu-
lar roles. Cell 76, 789–791.
18 Anastasiadis PZ & Reynolds AB (2000) The p120 cate-
nin family: complex roles in adhesion, signaling and
cancer. J Cell Sci 113, 1319–1334.
19 Riggleman BE, Wieschaus E & Schedl P (1989) Molecu-
lar analysis of the armadillo locus: uniformly distributed
transcripts and a protein with novel internal repeats are
associated with a Drosophila segment polarity gene.
Genes Dev 3, 96–113.
20 A Patapoutian LF Reichardt (2000) Roles of Wnt pro-
teins in neural development and maintenance. Curr Opin
Neurobiol 10, 392–399.
21 Deguchi M, Iizuka T, Hata Y, Nishimura W, Hirao K,
Yao I, Kawabe H & Takai Y (2000) PAPIN. A novel
multiple PSD-95 ⁄ Dlg-A ⁄ ZO-1 protein interacting with
neural plakophilin-related armadillo repeat pro-
tein ⁄ delta-catenin and p0071. J Biol Chem 275, 29875–
29880.
22 Lu Q, Paredes M, Medina M, Zhou J, Cavallo R, Peifer
M, Orecchio L & Kosik KS (1999) Delta-catenin, an
adhesive junction-associated protein which promotes cell
scattering. J Cell Biol 144, 519–532.
23 Wang B, Yang H, Liu YC, Jelinek T, Zhang L,
Ruoslahti E & Fu H (1999) Isolation of high-affinity
peptide antagonists of 14-3-3 proteins by phage display.
Biochemistry 38, 12499–12504.
24 Henriksson ML, Francis MS, Peden A, Aila M, Stefans-
son K, Palmer R, Aitken A & Hallberg B (2002) A non-

phosphorylated 14-3-3 binding motif on exoenzyme S
that is functional in vivo. Eur J Biochem 269, 4921–
4929.
25 Yeh TC, Ogawa W, Danielsen AG & Roth RA (1996)
Characterization and cloning of a 58 ⁄ 53-kDa substrate
of the insulin receptor tyrosine kinase. J Biol Chem 271,
2921–2928.
26 Alvarez CE, Sutcliffe JG & Thomas EA (2002) Novel
isoform of insulin receptor substrate p53 ⁄ p58 is gener-
ated by alternative splicing in the CRIB ⁄ SH3-binding
region. J Biol Chem 277, 24728–24734.
27 Obenauer JC, Cantley LC & Yaffe MB (2003) Scansite
2.0: Proteome-wide prediction of cell signaling interac-
tions using short sequence motifs. Nucleic Acids Res 31,
3635–3641.
28 Tzivion G & Avruch J (2002) 14-3-3 proteins: active
cofactors in cellular regulation by serine ⁄ threonine
phosphorylation. J Biol Chem 277, 3061–3064.
29 Obsil T, Ghirlando R, Anderson DE, Hickman AB &
Dyda F (2003) Two 14-3-3 binding motifs are required
for stable association of Forkhead transcription factor
FOXO4 with 14-3-3 proteins and inhibition of DNA
binding. Biochemistry 42, 15264–15272.
30 Zhou S, Liyanage U, Medina M, Ho C, Simmons AD,
Lovett M & Kosik KS (1997) Presenilin 1 interacts in
brain with a novel member of the Armadillo family.
NeuroReport 8, 1489–1494.
31 Yamagishi A, Masuda M, Ohki T, Onishi H & Mochi-
zuki N (2004) A novel actin bundling ⁄ filopodium-form-
ing domain conserved in insulin receptor tyrosine kinase

substrate p53 and missing in metastasis protein. J Biol
Chem 279, 14929–14936.
32 Thomas EA, Foye PE, Alvarez CE, Usui H & Sutcliffe
JG (2001) Insulin receptor substrate protein p53 locali-
zation in rats suggests a mechanism for specific polyglu-
tamine neurodegeneration. Neurosci Lett 309, 145–148.
33 Okamura-Oho Y, Miyashita T, Ohmi K & Yamada M
(1999) Dentatorubral-pallidoluysian atrophy protein
interacts through a proline-rich region near polygluta-
mine with the SH3 domain of an insulin receptor tyro-
sine kinase substrate. Hum Mol Genet 8, 947–957.
34 Dubois T, Kerai P, Zemlickova E, Howell S, Jackson
TR, Venkateswarlu K, Cullen PJ, Theibert AB, Larose
L, Roach PJ & Aitken A (2001) Casein kinase I associ-
ates with members of the centaurin-a family of phos-
phatidylinositol-3,4,5-trisphosphate-binding proteins.
J Biol Chem 276, 18757–18764.
35 Dubois T, Rommel C, Howell S, Steinhusen U, Soneji
Y, Moelling K & Aitken A (1997) 14-3-3 is phosphory-
lated by casein kinase I on residue 233. Phosphorylation
at this site in vivo regulates Raf ⁄ 14-3-3 interaction.
J Biol Chem 272, 28882–28888.
14-3-3 in neurodegenerative disease S. Mackie and A. Aitken
4210 FEBS Journal 272 (2005) 4202–4210 ª 2005 FEBS