Interactome mapping of the PhosphatidylInositol 3-Kinase-mammalian Target Of
Rapamycin pathway identifies Deformed Epidermal Autoregulatory Factor-1 as a new
Glycogen Synthase Kinase-3 interactor
Fanny Pilot-Storck1,6, Emilie Chopin1, Jean-François Rual3, Anais Baudot2, Pavel Dobrokhotov4,
Marc Robinson-Rechavi4, Christine Brun5, Michael E. Cusick3, David E. Hill3, Laurent
Schaeffer1, Marc Vidal3, Evelyne Goillot1*
1 UMR5239 Laboratoire de Biologie Moléculaire de la Cellule, Ecole Normale Supérieure de Lyon, 46
allée d’Italie, F-69007 Lyon, France
2 Spanish National Cancer Research Centre, C/ Melchor Fernández Almagro, 3, E-28029 Madrid, Spain
3 Center for Cancer Systems Biology and Department of Cancer Biology, Dana-Farber Cancer Institute,
and Department of Genetics, Harvard Medical School, 44 Binney Street, Boston, MA 02115, USA
4 Université de Lausanne, Département d'Ecologie et d'Evolution, Quartier Sorge, Lausanne, Switzerland.
5 TAGC Inserm U928, Université de la Méditerranée, Parc Technologique de Luminy, Case 928, F-13009
6 Present address : UMR955 INRA-ENVA, Ecole Nationale Vétérinaire d’Alfort, 7 avenue du Général de
Gaulle, F-94700 Maisons-Alfort, France
* Corresponding author: Evelyne Goillot, UMR5239 Laboratoire de Biologie Moléculaire de la Cellule,
Ecole Normale Supérieure de Lyon, 46 allée d’Italie, F-69007 Lyon, France. Tel : 33 472 72 81 24. Fax :
33 472 72 80 80. E-mail : email@example.com
Running title: PI3K-mTOR interactome
MCP Papers in Press. Published on April 5, 2010 as Manuscript M900568-MCP200
Copyright 2010 by The American Society for Biochemistry and Molecular Biology, Inc.
ABI1: abl-interactor 1
AD: Activation domain
AKT: PKB Protein kinase B
AMPK: 5'-AMP-activated protein kinase
APPL1: adaptor protein containing pleckstrin homology domain, phosphotyrosine binding (PTB) domain and leucine
ARHGEF11: Rho guanine nucleotide exchange factor (GEF) 11
BHLHB2/DEC1: basic helix-loop-helix domain containing, class B, 2
CK2: casein kinase 2
CKIP-1/ PLEKHO1: pleckstrin homology domain containing, family O member 1
co-AP: co-Affinity Purification
CPE: carboxypeptidase E
DB: DNA binding domain
DEAF1/ NUDR: Deformed epidermal autoregulatory factor-1
DMEM: Dulbecco's Modified Eagle Medium
EAAT4: excitatory amino acid transporter 4
FACS: Fluorescence Activated Cell Sorting
GβL: G protein beta subunit like
GLUT4: glucose transporter 4
GSK3A and B: glycogen synthase kinase 3 A and B
GTPases: guanosine triphosphatase
HIS3: Imidazoleglycerol-phosphate dehydratase
HSP: Hereditary Spastic Paralegia
5-HT: 5-HydroxyTryptamine, or serotonin
5-HT1A: Serotonin receptor 1A
HTTP: High throughput
hVps15: human vacuolar protein sorting 15 or phosphoinositide-3-kinase, class 3, regulatory subunit
hVps34: human vacuolar protein sorting 34 or phosphoinositide-3-kinase, class 3, catalytic subunit
IGF1: insulin-like growth factor 1
IGF1R: insulin-like growth factor I receptor
IP: Imuno precipitation
IR: Insulin receptor
IRS1, 2, 4: insulin receptor substrate 1, 2, 4
KIF1C: kinesin family member 1C
lacZ: beta-D-galactosidase LacZ
LZTS2: leucine zipper, putative tumor suppressor 2
MAP kinases/ MAPK: mitogen-activated protein kinases
MAPK14: mitogen-activated protein kinase 14
MDD: Major Depression Disease
MO25: calcium binding protein 39
MTM1: myotubularin 1
mTOR: mammalian target of rapamycin
mTORC1/2: mammalian target of rapamycin complex 1 and 2
NF-κB: nuclear factor kappa-B, subunit 1
PDK1: 3-phosphoinositide-dependent protein kinase
PI3,4,5-P3 : Phosphatidylinositol (3,4,5)-trisphosphate
PIKE: phosphoinositide 3-kinase enhancer
PIP: phosphatidylinositide phosphate
PKCζ: protein kinase C ζ
PPI: Protein-Protein Interaction
PRKAA1: 5'-AMP-activated protein kinase, catalytic alpha-1 chain
PTEN: phosphatase and tensin homolog
RC3H1/Roquin: ring finger and CCCH-type zinc finger domains 1
RHEB: Ras homolog enriched in brain
RHEBL1: Ras homolog enriched in brain like 1
RICTOR: RPTOR independent companion of MTOR
S6K1 and 2: ribosomal protein S6 kinase 1 and 2
SGK1: serum/glucocorticoid regulated kinase 1
SHIP1 and 2: SH2 containing inositol phosphatase 1 and 2
SPG21/Maspardin: spastic paraplegia 21
STK11/LKB1: serine/threonine kinase 11/
STRAD: STE20-related adapter protein
TNFR1: tumor necrosis factor receptor superfamily, member 1
TRAF2: TNF receptor-associated factor 2
TSC1: tuberous sclerosis 1
TSC2: tuberous sclerosis 2
URA3: Orotidine-5'-phosphate (OMP) decarboxylase
Y2H: Yeast two hybrid
ZBED3: zinc finger, BED domain containing 3
The PhosphatidylInositol 3-Kinase-mammalian Target Of Rapamycin (PI3K-mTOR) pathway
plays pivotal roles in cell survival, growth and proliferation downstream to growth factors. Its
perturbations are associated with cancer progression, type 2 diabetes and neurological disorders.
To better understand the mechanisms of action and regulation of this pathway, we initiated a
large scale yeast two-hybrid screens for 33 components of the PI3K-mTOR pathway.
Identification of 67 new interactions was followed by validation by co-affinity purification and
exhaustive literature curation of existing information. We provide a nearly complete, functionally
annotated interactome of 802 interactions for the PI3K-mTOR pathway. Our screen revealed a
predominant place for the Glycogen Synthase Kinases-3 (GSK3) A and B and the AMP-activated
protein kinase. In particular, we identified the Deformed Epidermal Autoregulatory Factor-1
(DEAF1) transcription factor as an interactor and in vitro substrate of GSK3A and GSK3B.
Moreover, GSK3 inhibitors increased DEAF1 transcriptional activity on the 5HT1A serotonin
receptor promoter. We propose that DEAF1 may represent a therapeutic target of lithium and
other GSK3 inhibitors used in bipolar disease and depression.
The PhosphatidylInositol 3-Kinases (PI3Ks) are a conserved family of lipid kinases that
phosphorylate phosphatidylinositol (PI) and phosphoinositides 3'-hydroxyl group, forming
second messengers. Class IA PI3K (hereafter “PI3K”) is activated by growth factor receptor
tyrosine kinases and upon stimulation by insulin or IGF1, PI3K triggers the formation of
PI3,4,5-P3 that recruits and activates kinases such as AKT and PDK1, mediating most effects
of insulin and IGF1 on cell metabolism, growth, proliferation and differentiation (1, 2).
Downstream of AKT, the mammalian Target Of Rapamycin (mTOR) kinase is an essential
activator of protein synthesis, promoting cell growth and proliferation (1, 3, 4). mTOR is
regulated by growth factors through AKT, by energy availability through the AMP-activated
kinase (AMPK) and by amino acid content through class III PI3K. Glycogen Synthase Kinase-
3 (GSK3) is another major target of the PI3K pathway, and its inhibitory phosphorylation by
AKT relieves its negative impact on cell cycle progression and cell growth (5). PI3K-mTOR
pathway is central for cell metabolism and proliferation and its perturbation is implicated in
many human diseases (1, 3, 4, 6). Mutations leading to PI3K-mTOR pathway activation are
important steps in the initiation and progression of tumours, and are frequently encountered in
human cancers. In contrast, down-regulation of PI3K pathway impairs cell responses to
insulin, leading to type 2 diabetes. Perturbations of the PI3K-mTOR pathways are also linked
to muscle atrophy, auto-immune and cardiovascular diseases. Moreover, GSK3 dysregulation
is associated with mood disorders and Alzheimer’s disease (7). Several components of the
PI3K-mTOR pathway are promising targets for anti-tumoral, metabolic and neurological
therapies (3, 4, 7).
The complexity of the PI3K-mTOR pathway necessitates innovative strategies to identify its
exact involvement in physiology and pathology and to predict the consequences of its
manipulation in therapy. A better comprehension of cell responses to PI3K-mTOR pathway
activation may come from the identification of new regulators or effectors of this pathway, and
this goal can now be reached via high-throughput approaches (8-11). We conducted a large-
scale yeast two-hybrid screen of 33 components of the PI3K-mTOR pathway. The resulting
interactions were supplemented with a manually curated set of literature interactions, providing
a comprehensive and annotated interactome for the PI3K-mTOR pathway. Our screen revealed
a predominant place for GSK3A, GSK3B and AMPK, and highlights their role in cancer,
metabolic diseases, immune response and neurological disorders. In particular, we
characterized a functional interaction of GSK3A and GSK3B with DEAF1 transcription factor
in the serotonergic pathway.
Cloning, yeast two-hybrid (Y2H) screens and co-Affinity Purification (Co-AP) experiments
Detailed description of cloning, Y2H screens and co-AP experiments are available in
Briefly, full-length ORF for our baits were cloned using the Gateway technology as DB- and AD-
expression vectors and transformed in MaV203 and MaV103 yeast strains. DB-expression
vectors were used for screening with a AD-cDNA library from E10.5 mouse embryo and both
DB- and AD-expression vectors were used for screening with the hORFeome1.1 library as
described previously (9, 12, 13). The activation of three reporter genes (HIS3, URA3 and lacZ)
was assessed in comparison to controls and an interaction was said positive if at least two
reporter genes were activated. For the cDNA screen, after selection of in-frame cDNAs,
interactions were retested using the gap repair technique (14). The gap repair technique was also
used to test interactions with GSK3A/GSK3B and RHEB/RHEBL1 constructs. A subset of 40
interactions was further tested in HEK293T cells, using full-length ORFs cloned as GST-bait and
Myc-prey vectors using the Gateway technology. Experiments were essentially performed as
described in (9).
For each selected PI3K-mTOR component, BIND (15), MINT (16), HPRD (17), APID (18) and
Pubmed databases were investigated for protein-protein interactions (PPI). Each PPI was verified
in the corresponding paper to check the exact reference of the protein and the technique. Only
binary interactions were retained, corresponding to the followings techniques: yeast or
mammalian two-hybrid experiments, in vitro kinase or other enzymatic assays with relevant
controls, in vitro binding of recombinant proteins purified from mammalian cell-free systems,
binding of a protein purified from mammalian cell-free systems to a membrane-immobilized
protein, crystallography and surface plasmon resonance analysis. A careful attention was
accorded to check that the interaction could not be indirect, due to a third component. Co-
immunoprecipitation and GST-pull down performed in vivo were not retained. However, in vitro
kinase assays involving mTOR or STK11/LKB1 kinases were often performed with kinases
isolated from cells as a complex (mTORC1/2 complex for mTOR; in complex with MO25 and
STRAD for STK11). If controls for these specific kinase activities were suitable, interactions
were included with the corresponding indication. Interactions involving a purified protein for
which the exact isoform could not be determined were not retained. Described interactions
mostly involve human proteins but sometimes involve proteins from mouse or other mammals.
For each interaction, the PMID reference referring to the paper describing the binary interaction
Functional annotation of each interactor for its molecular and sub-cellular functions and its
pathway involvement was deduced from NCBI-Gene and Pubmed data banks. Proteins were
classified according to a limited number of chosen terms rather than to Gene Ontology
annotations that were either too much detailed or not accurate enough according to the considered
protein. As a protein can be engaged in several PPIs with different components of the PI3K-
mTOR pathway, interactors may be redundant whereas interactions are not. Therefore, the
annotation of interactors is formally referred to as the annotation of interactions.
Kinase assay with GSK3, MLK3 and DEAF1 constructs
- Protein purification
GST-GSK3A, GST-GSK3B, GST-GSK3A-KA, GST-GSK3B-KA Gateway vectors were
transfected in HEK293T cells as described for co-AP experiments. Cells were lysed with co-AP
lysis buffer for 30min on ice. Lysates were sonicated and pre-cleared by centrifugation for 10min
at 14,000rpm at 4°C. Pre-cleared lysates were incubated with Immobilized Glutathione beads
(Perbio) for 1h at 4°C. Beads were washed extensively three times with lysis buffer. GST fusion
proteins were eluted with PBS, 50mM glutathione (Perbio) for 15min at room temperature.
Aliquots of eluates were run on a polyacrylamide/SDS gel in parallel to a BSA scale. The gel was
stained with a Coomassie solution and the quantity of GST fusion proteins was estimated.
Adjusted amounts of proteins were then analyzed by western-blot using anti-GST antibody
(figure 6B; left gel). Myc-DEAF1, Myc-DEAF1-1, Myc-DeaF1-2, Myc-DEAF1-3 Gateway and
pCDN3-M2-MLK3 (kindly provided by K. A. Gallo) vectors were transfected in HEK293T cells
as described for co-AP experiments. Each well of 6-well plates was lysed with 250μL of IP
buffer 1 (300mM NaCl, 50mM Tris pH 7.4, 0,5mM EDTA, 1% Triton X-100, Complete protease
inhibitor cocktail (Roche)) for 30min on ice. Lysates were sonicated, pre-cleared by
centrifugation for 10min at 14,000rpm at 4°C. 250μL of IP buffer 2 (50mM Tris pH 7.4, 0,5mM
EDTA, 1% Triton X-100, Complete protease inhibitor cocktail (Roche)) and mouse anti-Myc
9E10 antibody (Covance) or anti-M2 flag (Sigma) were added to the lysates for overnight
agitation at 4°C. Protein A- and G-sepharose beads were added for 1h at 4°C and then
extensively washed with IP buffer 3 (150mM NaCl, 50mM Tris pH 7.4, 0,5mM EDTA, 1%
Triton X-100, Complete protease inhibitor cocktail (Roche)). All tree IP buffers were
supplemented with phosphatase inhibitors (200μM NaF, 1mM sodium orthovanadate, 100μM β-
glycerophosphate). Finally, beads were resuspended in 1X kinase buffer (25mM Hepes pH 7.4,
10mM MgCl2) and 10μL of beads suspensions were run on a polyacrylamide/SDS gel,
transferred to PVDF membranes and analyzed by western blotting with anti-Myc (figure 6C) and
M2 (figure 6B, right gel) antibodies to estimate the quantity of Myc-DEAF1 and M2-MLK3
- Kinase assay
Similar amounts of GST-GSK3 proteins (around 70ng) were added to Myc-DEAF1 or M2-
MLK3 recombinant proteins (around 50ng). The mix was adjusted to final concentrations of
25mM Hepes pH 7.4 and 10mM MgCl2. 30μCi of 32-P γ-ATP were added to the mix. After
incubation for 30min at 30°C, the reaction was stopped by addition of 5X loading buffer and
denaturation at 95°C for 5min.
Aliquots run on polyacrylamide/SDS gels and transferred on PVDF membranes that were
revealed using a PhosphorImager system (Fuji) for radioactivity measures.
- Quantification of Myc-Deaf1 constructs
The membranes were then subjected to western-blotting with mouse anti-Myc (9E10) and goat
anti-mouse IRDye 800 secondary antibody (LI-COR) to reveal and quantify Myc-DEAF1
proteins on Odyssey LI-COR scanner (ScienceTec).
Adjusted radioactive signal from Myc-DEAF1 proteins was calculated by the ratio radioactive
RT-PCRs for 5HT1A and DEAF1
mRNAs were extracted from RN46A and HEK293T cells using NucleoSpin RNA II kit
(Macherey-Nagel). 500ng of gel-verified RNA were retro-transcribed using RevertAid H Minus
M-Mul V RT (Fermentas). DNA from RN46A cells was extracted by lysis in 5mM EDTA, 5 mM
Tris pH 8, 100mM NaCl, 0,2% SDS and 0.3μg/μL of proteinase K, followed by precipitation.
PCR were performed using Qiagen Taq polymerase on the retro-transcribed products to check the
expression of actin, 5HT1A and Deaf1 in RN46A cells and of actin and DEAF1 in HEK293T
cells. Couples of oligos used for 5HT1A amplification were the followings: sense and antisense 5-
HT1A oligos described in (19), sense 5-HT1A (19) plus 5-HT1A new antisense oligo
(GCGGTGCCGACGAAGTTC), and 5HT1A-FOR (CCGCACGCTTCCGAATC) plus 5HT1A-
REV (ACCTGGCTGTCCGTTCAG) oligos. Oligos used for Deaf1 amplification were the
followings: rDeaf1-FOR (TGCACCTGTGCTGCCTGTTG) and rDeaf1-REV
(CGATCTGGCAGCTGTCCTGAT) (on rat RN46A products), hDEAF1-FOR
(GAACGCGGCATCCATCTCAG) and hDEAF1-REV (CTTGCGTTGGCAGAAGGTGG) (on
human HEK293T products). PCR for actin and Deaf1 were positive on all tested RT products, as
well as PCR for 5HT1A on RN46A cells DNA (data not shown). Identity of RT-PCR products
for Deaf1 from RN46A and 293T cells was assessed by sequencing.
The luciferase plasmid 5-HT1A(C) was kindly provided by Dr Paul R. Albert’s lab and is
described in Lemonde et al (19). GST, Deaf1-Myc and Myc Gateway vectors are described in
previous parts. pCMV-SPORT-βgal (Invitrogen) and pGFP-S6 (20) plasmids were used as
reporter plasmids for normalization.
- Transfections and drug treatment
HEK293T cells were grown at 37°C, 5% CO2 in DMEM supplemented with 10% fetal bovine
serum. Cells were seeded in 12-well plates and transfected using calcium chloride. Each well was
transfected with 200ng of 5-HT1A(C) vector, 50ng of pCMV-SPORT-βgal, 150ng of GST,
Deaf1-Myc or Myc vector and 400ng of pcDNA3 to complete the DNA mix. Drugs were added
at transfection. Lithium and sodium chloride were added at a final concentration of 5mM.
Azakenpaullone (Calbiochem) was diluted in DMSO and added at a final concentration of 1μM
or 500nM. An equivalent volume of DMSO (0,25μL) was added in control wells. Medium was
changed 12h after transfection and drugs were replaced at the same concentrations.
RN46A cells (kindly provided by Dr Scott Whittemore’s laboratory) were grown at 33°C, 5%
CO2 in Neurobasal medium (Invitrogen) supplemented with 10% fetal bovine serum and
glutamine. Cells were seeded in 12-well-plates. The transfection mix was composed of 100μL of
OptiMEM (Invitrogen), 4μL of Fugene HD (Roche), 200ng of 5HT1A(C) vector, 50ng of CMV-
GFP vector, 150ng of GST, Deaf1-Myc or Myc vector and 400ng of pcDNA3. Drugs were added
in the same conditions as for HEK293T cells, except that that the medium was not changed after
- Luciferase assay
Cells were harvested 36h after transfection. HEK293T cells were lysed with 250μL of Passive
Lysis Buffer (Promega). 10μL of lysate were used for luciferase activity detection using the
Luciferase Assay System (Promega). 50μL of lysate were used for β-galactosidase activity
detection and were added to 50μL of detection buffer containing 1mM MgCl2, 50μM β-
mercaptoethanol, 40μg ONPG. After incubation at 37°C, the plate was read at 405nm.
RN46A cells were dissociated with Versene solution (Invitrogen) and resuspended in 1ml of
PBS. 200μL were used for FACS analysis (Becton-Dickinson) of GFP transfected cells. The
remaining cells were lysed with 100μL of Reporter Lysis Buffer (Promega). 10μL of lysate were
used for luciferase activity detection using the Luciferase Assay System (Promega). 10μL of
lysate were used for protein assay using Bio-Rad Protein Assay.
- Statistical analysis
Independent experiments were performed at least 6 times. Adjusted luciferase activity was
corrected for transfection efficiency and cell density by calculating the ratio of luciferase
activity/β-galactosidase activity for HEK293T cells. As β-galactosidase activity detection was
not sensitive enough for normalization in RN46A cells (data not shown), adjusted luciferase
activity was calculated as the ratio of luciferase activity/GFP transfection rate/protein
concentration. Statistical analysis was performed using Student t-test.
Cloning of the PI3K-mTOR pathway components
PI3K-mTOR pathway is multi-modular, each functional module integrating various inputs for
downstream effectors (1, 2). Thirty-seven components, considered as essential mediators
involved in IGF1 signaling were selected (Figure 1). They are schematized as part of a
receptor module (IGF1, IGF1R, IR, IRS1, IRS2, IRS4), a PI3K module (p85α, p110α,
hVps34, hVps15, SHIP1, SHIP2, PTEN, PDK1, AKT1, AKT2, AKT3, PIKE, GSK3A,
GSK3B), an mTOR module (mTOR, GβL, RICTOR, raptor, TSC1, TSC2, RHEB, RHEBL1,
S6K1, S6K2, LKB1, AMPK), or a related proteins module (PKCζ, SGK1, CK2, CKIP1,
MTM1). The functional paralogy for RHEB and RHEBL1 or GSK3A and GSK3B is indicated.
Description of all components is provided (Supplementary data 1).
Selected components were cloned by the Gateway recombinational cloning into yeast two-hybrid
(Y2H) destination vectors to generate GAL4 DNA binding domain (DB)- or activation domain
(AD)- fusion proteins (9, 21). IR, IRS2, IRS4 and SHIP1 failed to clone, leaving 33 constructs for
Yeast two-hybrid screens and co-affinity purification experiments
As the PI3K-mTOR pathway is conserved and ubiquitously expressed, it was important to screen
using the broadest possible libraries. Two stringent Y2H screens were carried out, one with an
E10.5 mouse embryo library and the other with the human ORFeome v1.1 ORF library (9, 21).
Both screens followed optimized protocols in which high quality parameters are assessed (14, 22,
In total, the cDNA and the ORFeome screens identified 68 and 8 protein-protein interactions
(PPIs) respectively, involving 15 PI3K-mTOR pathway components (Figure 2, Supplemental
Table I). Two PPIs were found in both screens, showing the complementarity of these screens.
Three PPIs have been described to occur in the literature and four other interactions found with
GSK3A had been previously characterized with its close paralog GSK3B, giving 67/74 (91%)
new potential interactions (Supplemental Table II). There were 35 PPIs (47%) found at least
twice independently (Figure 2, Supplemental Table I).
GSK3A and PRKAA1, the AMPK α1 catalytic subunit, were the most connected proteins, with
17 and 27 PPIs respectively.
Paralogs in the screens had sharply different numbers of PPIs. We found four PPIs for RHEBL1
but none for RHEB, and 17 PPIs for GSK3A but 2 for GSK3B. Using the gap-repair technique
(14), we tested the interaction between RHEBL1 interactors and RHEB. Y2H interactions were
retrieved with RHEB and RHEBL1 only when their last amino acids that constitute a prenylation
motif (24) were removed (Supplemental Figure 1). In this case, RHEB interacted with 3 out of 4
interactors of RHEBL1 (Supplemental Figure 1, Figure 2, Supplemental Table I).
To validate Y2H interactions, we tested half (N=40) of the PPIs by co-affinity purification (co-
AP) assays in human HEK293T cells (9). Transfected proteins were expressed in 33 co-APs and
nineteen co-APs (58%) tested positive (Supplemental Figure 2, Supplemental Table I), in line
with expectation from previous such validation efforts (25, 26).
The protein interactions reported in this publication have been submitted to the IMEx
(http://imex.sf.net) consortium through IntAct (PMID: 17145710) and assigned the identifier IM-
We manually curated from literature all binary interactions for each PI3K-mTOR pathway
component. Verified interactions were added to our new interactions in a resulting “literature-
completed interactome” (LC-interactome) containing 802 distinct interactions (Supplemental
Table II). The interactors were annotated for their molecular function and their sub-cellular
functions, based on validated data from literature (Supplemental Table II, Figure 3).
Accordingly, 648 interactors were annotated for their participation to particular signaling
pathways (Supplemental Table II, Figure 4).
GSK3A and GSK3B protein-protein interactions
The GSK3A and GSK3B paralogs have high sequence similarity and almost identical kinase
domains (27), but have divergent N-terminal regions, spanning from amino acid (aa) 1 to 90 for
GSK3A and 1 to 27 for GSK3B (Figure 5A). To question if GSK3A PPIs involved its specific
N-terminal domain or the homologous part shared with GSK3B, we made the following
constructs: full-length GSK3A (GSK3A-wt) and GSK3B (GSK3B-wt), N-GSK3A (aa 1-90) and
C-GSK3A (aa 91-483) (Figure 5A). Two additional constructs, GSK3A-SA and GSK3B-SA,
had S21 and S9 respectively mutated to alanine to render them constitutionally active kinases (28,
29). These constructs can be compared with the initial GSK3A construct used in our screens
(GSK3AΔ), which has an in-frame deletion of 31 amino acids from aa 7 to 38 removing the
negative Ser21 regulatory site (27) (Figure 5A). These 6 constructs were screened against 16
Y2H interactors of GSK3AΔ (Figure 5B). GSK3A-SA behaved as GSK3AΔ, interacting with all
tested interactors, and GSK3A-wt exhibited 15 interactions. N-GSK3A did not interact with any
GSK3AΔ interactor, indicating that these PPIs did not involve the GSK3A specific N-terminal
domain. In contrast, C-GSK3A interacted with 14 GSK3AΔ interactors. GSK3B-SA interacted
with 15 over the 16 interactors whereas GSK3B-wt exhibited 10 interactions. In conclusion, all
tested interactions involved the homologous parts of GSK3A and GSK3B. Constitutive activation
of these kinases facilitated Y2H interaction, especially for GSK3B.
To test if the kinase activity of GSK3A and GSK3B was necessary for these Y2H interactions,
we cloned the kinase-dead forms GSK3A-KA and GSK3B-KA, in which K148 and K85 were
respectively replaced by an alanine (30, 31). To prevent a deleterious phosphorylation on S21 and
S9, we also constructed GSK3A-SA-KA and GSK3B-SA-KA forms respectively, combining
both described mutations. The KA mutations abolished most interactions for both kinases,
independently of the presence of the inactivation site. There was only one interactor for GSK3B-
KA and two for GSK3A-KA (Figure 5B).
We tested seven of these interactions by co-AP in HEK293T cells (Figure 5C, Supplemental
Table I). GSK3A-SA and GSK3B-SA interacted with all tested GSK3AΔ interactors, GSK3A-wt
and GSK3B-wt with 6 and 4 GSK3AΔ interactors respectively, and GSK3A-KA and GSK3B-KA
with 3 and 4 GSK3AΔ interactors respectively. Seemingly, the co-AP assay was less sensitive to
GSK3 kinase activity than Y2H.
In conclusion, all GSK3AΔ tested interactors were shared by GSK3B and were favored by
constitutive activation of the kinases, whereas kinase inactivation abolished most interactions.
Different sets of interactors were observed for wild-type, constitutively active and inactive forms
of GSK3A and GSK3B (Figure 5B-C), suggesting specificities for these two kinases.
DEAF1 is a new GSK3 substrate
Our Y2H screens identified the Deformed Epidermal Autoregulatory Factor 1 (DEAF1, also
called NUDR) transcription factor as a new interactor for GSK3A and GSK3B and these
interactions were confirmed by co-AP assay in human cells (Figures 2 and 5, Supplemental
Figure 2, Supplemental Table I).
We tested whether DEAF1 was a substrate for GSK3A and GSK3B kinase activity. GSK3A or
GSK3B mutated for kinase activity showed impaired Y2H interactions with DEAF1 (Figure 5B).
DEAF1 has 14 putative conserved sites for GSK3 phosphorylation (S/T XXX S/T sequences,
where S/T should be primed by another kinase (32)), with a stretch of 7 sites from aa 328 to 358,
a region also present in the mouse DEAF1 cDNA isolated in the cDNA Y2H screen (aa 195 to
565) (Figure 6A). Several Myc-DEAF1 fusion proteins were tested for their capacities to be
phosphorylated by GSK3A and GSK3B: full-length DEAF1 (DEAF1-FL, aa 1 to 565), DEAF1-1
(aa 1 to 194), DEAF1-2 (aa 195 to 405) and DEAF1-3 (aa 406 to 565) (Figure 6A). Wild-type
and kinase-dead GSK3A and GSK3B kinases were purified and used in in vitro kinase assays
(Figure 6B-C). MLK3, a recently identified GSK3 substrate ((33) and our screen) was used as
positive control (Figure 6B-C). DEAF1-FL, DEAF1-1 and DEAF1-2 proteins were
phosphorylated by wild type but not kinase-dead forms of GSK3A and GSK3B (Figure 6C). The
DEAF1-FL and DEAF1-2 proteins were phosphorylated to a similar extent, whereas DEAF1-1
showed a little more phosphorylation (Figure 6C-D). DEAF1-3 phosphorylation levels remained
at background in all conditions (Figure 6C-D). DEAF1 likely represents a substrate for GSK3A
and GSK3B kinases, with several phosphorylation sites distributed from aa 1 to 405.
DEAF1 transcriptional activity upon 5HT1A promoter is increased by GSK3 inhibition
We next tested whether GSK3 kinase activity affected DEAF1 transcriptional activity. A major
transcriptional target for DEAF1 is the 5HT1A serotonin receptor gene (19, 34). We examined
DEAF1 transcriptional activity upon GSK3 inhibition with a luciferase reporter under control of
5HT1A promoter (19). Transfection of the full-length Myc-DEAF1 construct into HEK293T cells
caused luciferase activity to increase (Figure 7A, p<0.005). Both, lithium and azakenpaullone,
two inhibitors of GSK3 acting through different mechanisms (35, 36), significantly increased
Myc-DEAF1 activity as reflected by increased luciferase activity in HEK293T cells (Figure 7A,
p<0.001). Similarly, Myc-DEAF1 transfection into rat RN46A serotonergic cells markedly
increased luciferase activity (Figure 7B, p<0.005), and treatment by lithium or azakenpaullone
stimulated Myc-DEAF1 activation of luciferase activity. Whereas lithium stimulated in RN46A
cells as strongly as in HEK293T cells (p<0.001), azakenpaullone only moderately stimulated
luciferase activity (p<0.05 at 1µM, p<0.025 at 500nM).
DEAF1 is broadly expressed (37) and RT-PCR experiments confirmed that DEAF1 was
expressed in HEK293T and RN46A cells (data not shown: Material and Methods). We tested the
effect of GSK3 inhibition upon endogenous DEAF1 activity. In both cell lines, luciferase activity
under control of 5HT1A promoter was significantly increased by GSK3 inhibitors, with the
strongest stimulation by lithium in RN46A cells (Figure 7C-D, p<0.001).
In summary, DEAF1 activation of 5HT1A promoter activity increased upon GSK3 inhibition in
HEK293T and RN46A cells, suggesting an inhibitory role of GSK3 phosphorylation on DEAF1.
Protein-protein interaction mapping for the PI3K-mTOR pathway
Y2H mapping for 33 components of the PI3K-mTOR pathway against two different libraries
identified 74 interactions, 67 of which have not been described previously. This screen revealed
two highly connected proteins: PRKAA1, the AMPK α1 catalytic subunit (27 interactors) and
GSK3A (17 interactors). Literature curation provided a set of 802 functionally annotated PPIs for
the PI3K-mTOR pathway (LC-interactome) which offers a comprehensive picture of the
connectivity and the predominant cellular processes involving the PI3K-mTOR pathway
(Supplementary data 2). The low overlap between our screen and the interactions from
literature (7/74) may be explained by two parameters: sampling sensitivity and assay sensitivity.
Sampling sensitivity refers to the interactions that can be identified in a single trial of an assay.
Sampling sensitivity was estimated at 45% in a comparable Y2H screen (23), and 6 screens were
needed to reach 90% saturation. Successive iterations of our screen would likely achieve a higher
degree of saturation. In parallel, assay sensitivity designates the fraction of all interactions that
can be identified by an assay with its specific experimental conditions. In parallel to the high
stringency of the Y2H criteria used in this study (low copy plasmids, two reporters activated), its
intrinsic limitations may include the cDNA and ORF isoforms and level of representation in the
libraries used, the activation status and the post-translational modifications of the baits in yeast,
that may be different from the ones encountered in previous screens. Moreover, full-length baits,
instead of domains, may have prevented some PPI (26) and lipid interacting domains present in
several components of the PI3K-mTOR pathway may have impaired the nuclear translocation
required in Y2H screening. Interestingly, assay sensitivity for different methods of PPI
identification was estimated to a similar value around 20-35% and our screen probably lies in the
same range (23). Altogether, sampling and assay sensitivities may explain the complementarities
of the different approaches used in PPI identification for PI3K-mTOR components.
The new discovered interactions from this study provide insights on GSK3A and GSK3B
specificities and suggest new paths for the comprehension of PI3K-mTOR pathway involvement
in cancer, metabolism and diabetes, immune response and neurobiology.
Implications for GSK3A and GSK3B specificities
GSK3A and GSK3B kinases are highly related paralog kinases that share a large redundancy,
although recent studies revealed specific functions for each of them (27, 30, 38-41).
Understanding isoform redundancies and specificities represents an important stake for cell
biology and drug development (42). However, GSK3A received little attention compared to
GSK3B. Our screens identified 17 common interactors for GSK3A and GSK3B, reinforcing a
functional redundancy between them. Y2H interactions were nearly abrogated when kinase-dead
forms of GSK3A and GSK3B were used, indicating that these interactors may be substrates for
However all interactors did not interact with the same strength and with the equivalent constructs
for each kinase. For example, C14orf129 interacted with C-GSK3A and constitutively active
GSK3A but with kinase-dead GSK3B in Y2H. These different affinities for GSK3A and GSK3B,
depending on their activation status, may lead in vivo to a preferential interaction with one of
these two kinases or to a sequestration by one of them. Beyond an apparent redundancy, the
status of GSK3A and GSK3B for their kinase activity may represent a determinant element of
their specificities. Characterization of these respective interactions could help understanding the
coordinated activities of these kinases and developing focused strategies for targeting specific
PI3K-mTOR pathway and cancer
Many components of the PI3K-mTOR pathway act as oncogenes or tumour suppressor genes and
mutations leading to an activation of the PI3K-mTOR pathway may account for 30% of human
cancers (43). Many drugs targeting PI3K, AKT or mTOR are currently tested for cancer therapy
(42). Several interactions discovered in our Y2H may participate to specific aspects of tumour
initiation or progression.
Our screen identified a new interaction between TSC1 and MAPK14 (p38α), one of the four
members of p38 MAP kinase family. p38 MAP kinases regulate inflammatory and stress
response and behave as tumour suppressors (44). The TSC1-TSC2 complex inhibits mTOR, a
translational activator promoting cell growth and proliferation, and mutations in TSC1 or TSC2
lead to a tumour syndrome (45). p38 may regulate mTOR activation through TSC1
New interactions may link the PI3K-mTOR pathway to the Wnt-beta-catenin pathway, which is
also involved in tumorigenesis (46). GSK3 is at the crossroad of these two signaling pathways.
Upon activation of Wnt signaling pathway, GSK3, in complex with Axin, phosphorylates beta-
catenin and targets it to degradation (32). In our Y2H screens, GSK3A and GSK3B interacted
with LZTS2 and ZBED3. LZTS2 tumour suppressor promotes beta-catenin nuclear export,
impairing its transcriptional activity (47). ZBED3, an Axin-binding protein, inhibits beta-catenin
phosphorylation by GSK3B (48). The interactions between GSK3 and LZTS2 and ZBED3
suggest new coordination for beta-catenin inhibition, ensuring a precise control of this potent
AMPK, the “cell energy sensor”, inhibits mTOR through the activation of the TSC1-TSC2
complex (49) and AMPK activating drugs are candidate for cancer therapies (50). Interestingly,
LZTS2 was also found as a new interactor for PRKAA1, suggesting a putative coordination
between GSK3 and AMPK in beta-catenin regulation. A similar link between these two kinases
already exists in sequential phosphorylation and activation of TSC2 (51). This interaction
between AMPK and LZTS2 may reveal another role for AMPK in cancer.
Moreover, AMPK is involved in the control of epithelial polarity, actin organization and cell
proliferation downstream to LKB1 (52-54). We found a new interaction between PRKAA1 and
ABI1, the Abelson interactor protein-1. ABI1 regulates actin polymerization and its expression
correlates with cell migration and invasiveness of cancer cell lines (55, 56). ABI1 may thus
mediate part of AMPK control of cell polarity and tumour progression.
Finally, we describe a new interaction between PRKAA1 and KIF1C, a plus-end microtubule
motor that promotes podosome dynamics (57). KIF1C expression is highly predictive of brain
metastasis in lung cancer (58) and immunotherapy against KIF1C appears promising in glioma
(59). We suggest that KIF1C interaction with AMPK may modulate its activity in cancer
PI3K-mTOR pathway, metabolism and diabetes
AMPK is activated upon high AMP/ATP ratio and turns on ATP-generating processes while
switching off ATP-consuming activities (60). AMPK activating drugs, such as AICAR or
metformin, are used in type 2 diabetes treatment. It was recently shown that AMPK regulates the
circadian clock of peripheral organs (61). We describe a new interaction between PRKAA1 and
BHLHB2, or DEC1, a transcription factor regulating the expression of metabolic and clock genes
which are required for circadian rhythm (62, 63). This interaction may participate to the control
of circadian rhythm by AMPK, likely adapting metabolism to animal activity.
Carboxypeptidase E, CPE, was another new interactor of PRKAA1. Mutation of CPE is
associated with type 2 diabetes (64-66). The regulation of CPE activity by AMPK could
participate to the therapeutic benefits of AMPK activation in type 2 diabetes.
GSK3 is another prime target for type 2 diabetes management (46). In our screen, GSK3A and
GSK3B interacted with ARHGEF11, or PDZ-RhoGEF, a small GTPases activator. ARHGEF11
polymorphism is associated with insulin resistance and type 2 diabetes (67-69). Several small
GTPases are implicated in the trafficking of glucose transporter 4 (GLUT4) following insulin
stimulation (70). We propose that GSK3 interaction with ARHGEF11 may regulate glucose
uptake in response to insulin through the modulation of GLUT4 transport.
Finally, we describe a new interaction between RHEB/RHEBL1 and APPL1, an adaptor protein
that mediates adiponectin-induced sensitization of the insulin pathway (71). APPL1 activates
AMPK, leading to S6K inhibition and subsequent release of IRS1 inhibition (see Figure 1) (72).
This likely occurs through RHEB inhibition as RHEB overexpression impairs APPL1 effect (72).
APPL1 direct interaction with RHEB and RHEBL1 may thus reinforce APPL1-mediated
sensitization of insulin signaling.
PI3K-mTOR pathway and immune response
The PI3K-mTOR pathway participates to many aspects of the immune response and rapamycin, a
mTOR inhibitor, is used as an immunosuppressive drug (73, 74). We found a new interaction
between RHEB/RHEBL1, two mTOR activators, and TRAF2, an adaptor protein that lies
downstream to TNFR1 and CD40 receptors and mediates the activation of the MAPK and NF-κB
pathways (75). TRAF2 participates to various aspects of the immune response (76, 77) and
RHEB negatively regulates MAPK activation through B-Raf inhibition (78). TRAF2 interaction
with RHEB and RHEBL1 may represent another possible level of control of immune responses
by PI3K downstream to TNFα and CD40 ligand.
In addition, we identified an interaction between PRKAA1 and RC3H1, or Roquin. In mouse,
Roquin mutation impairs the degradation of particular mRNAs, leading to the development of
severe auto-immunity (79, 80). AMPK may influence Roquin activity and modulate certain
aspects of auto-immunity.
PI3K-mTOR pathway and neurobiology
During development, GSK3 regulates neurite extension and neuronal architecture (5). In adult,
GSK3 has been implicated in specific behavioral responses (81). In our Y2H screens, GSK3A
and GSK3B interacted with ARHGEF11 which increases EAAT4 glutamate transporter activity
in brain and participates to neuronal morphogenesis (82, 83). GSK3 interaction with ARHGEF11
may explain part of GSK3 involvement in synaptic plasticity. More broadly, we suggest that this
interaction may participate to regulated trafficking in different contexts, such as the insulin
pathway (see above) and neurogenesis.
Our Y2H screens also found SPG21, or Maspardin, interacting with GSK3A. SPG21 mutation
leads to Hereditary Spastic Paralegia (HSP), a neurodegenerative disease associated with
dementia (84). Spastic disorders are also encountered in hereditary cases of Alzheimer’s disease
(85). Importantly, GSK3 is involved in the progression of Alzheimer’s disease and chemical
inhibition of GSK3 alleviates beta-amyloid accumulation (7). The interaction between GSK3A
and SPG21 may reveal a mechanistic proximity between HSP and Alzheimer’s disease.
Possible involvement of DEAF1-GSK3 interaction in mood disorders
Finally, we identified DEAF1 transcription factor as an interactor and an in vitro substrate for
GSK3A and GSK3B. GSK3 inhibitors increased DEAF1 transcriptional activity on the 5HT1A
promoter in both HEK293T and RN46A cells, suggesting an inhibitory regulation of DEAF1 by
Variations in serotonin neurotransmitter (5-HT) availability and signaling cascade are associated
with psychiatric disorders such as Major Depression Disease (MDD). Among several receptors,
the 5HT1A receptor for serotonin is a major target of antidepressant treatments (86-89). Pre-
synaptic 5HT1A receptors act as inhibitory autoreceptors in serotonergic neurons located in
midbrain raphe nuclei. On the contrary, post-synaptic 5HT1A receptors are found in 5-HT
responsive areas that are limbic regions and specific cortical layers.
In previous reports, DEAF1 repressed 5HT1A expression in HEK293T cells and in the
serotonergic cells RN46A that express 5HT1A, but enhanced 5HT1A expression in non-
serotonergic 5HT1A-expressing cells such as SN48, NG108-15 or SKN-SH cells (19, 34).
However, DEAF1 enhanced 5HT1A expression in several hippocampal and septal cells (19).
These data likely reflect an opposite regulation on presynaptic versus postsynaptic 5HT1A
expression (34) and perhaps a phenotypic derivation of the HEK293T cells and RN46A cells we
used. The RN46A cells we used did not express 5HT1A (data not shown) and may behave as non-
serotonergic cells. The stimulation of 5HT1A expression by DEAF1 that we observed in both cell
lines is in accordance with neither of them expressing 5HT1A.
A reduction of 5HT1A expression is observed in post-synaptic area (86, 90-93) and in raphe
nuclei (86, 90, 94, 95) during depression or bipolar disease (89). Accordingly, a blunted response
to 5HT1A activation has been associated with depression. Most antidepressants increase
serotonin transmission and lead to improved 5HT1A function in post-synaptic areas (86, 87, 89,
90). These data suggest that increasing 5HT1A expression should alleviate depression symptoms.
In parallel, lithium has long been used in treatment of bipolar disease and as an adjuvant to
antidepressant drugs (36, 81, 86). The preeminent mechanism of action of lithium in psychiatric
disorders is likely represented by GSK3 inhibition (7, 36). Accordingly, lithium behavioral
responses are increased by GSK3B haploinsufficiency and reproduced by another GSK3 inhibitor
(96) whereas GSK3B haploinsufficiency and GSK3B inhibition alleviates aberrant behaviors due
to 5-HT deficiency (97). Other mood stabilizing agents such as clozapine and valproate also
inhibit GSK3 activity (81, 98). Conversely, an increased GSK3B activity is observed in
prefrontal cortex of MDD subjects (99) whereas both 5HT1A and DEAF1 were decreased in
prefrontal cortex of depressed women, suggesting a positive regulation of 5HT1A transcription by
DEAF1 in these areas (93). Importantly, treatment with lithium or divalproex increased 5HT1A
expression in bipolar disorder patients (100). Our finding that GSK3 phosphorylates DEAF1 and
impairs DEAF1-driven expression of 5HT1A suggests that abnormal GSK3B activation reduces
5HT1A expression, contributing to depressive symptoms, whereas GSK3 inhibition by lithium or
another drug would increase DEAF1-driven 5HT1A expression in 5HT-responsive areas and
restore 5HT transmission. GSK3 and DEAF1 interaction might explain part of GSK3
involvement in bipolar disease and depression and DEAF1 may represent a therapeutic target of
lithium and other GSK3 inhibitors used in these disorders.
We are grateful to J. Woodgett, P. R. Albert, S. Whittemore, K. A. Gallo, J. J. Zhao, T. M.
Roberts, M. Billaud, C. Erneux, G. Thomas and K. Salehi-Ashtiani for sharing reagents and to P.
Lamesch, T. Hao and K. Gauthier for technical advice. This work was supported by ARC grant
N°3853 (Association de Recherche sur le Cancer) and F. P. S. by an AFM grant (Association
Française contre les Myopathies).
The authors declare that they have no conflict of interest.
1. Engelman, J. A., Luo, J., and Cantley, L. C. (2006) The evolution of phosphatidylinositol
3-kinases as regulators of growth and metabolism. Nat. Rev. Genet. 7, 606-619.
2. Taniguchi, C. M., Emanuelli, B., and Kahn, C. R. (2006) Critical nodes in signalling
pathways: insights into insulin action. Nat. Rev. Mol. Cell Biol. 7, 85-96.
3. Wullschleger, S., Loewith, R., and Hall, M. N. (2006) TOR signaling in growth and
metabolism. Cell 124, 471-484.
4. Shaw, R. J., and Cantley, L. C. (2006) Ras, PI(3)K and mTOR signalling controls tumour
cell growth. Nature 441, 424-430.
5. Jope, R. S., and Johnson, G. V. (2004) The glamour and gloom of glycogen synthase
kinase-3. Trends Biochem. Sci. 29, 95-102.
6. Wymann, M. P., and Marone, R. (2005) Phosphoinositide 3-kinase in disease: timing,
location, and scaffolding. Curr. Opin. Cell Biol. 17, 141-149.
7. Cohen, P., and Goedert, M. (2004) GSK3 inhibitors: development and therapeutic
potential. Nat. Rev. Drug Discov. 3, 479-487.
8. Stelzl, U., Worm, U., Lalowski, M., Haenig, C., Brembeck, F. H., Goehler, H.,
Stroedicke, M., Zenkner, M., Schoenherr, A., Koeppen, S., Timm, J., Mintzlaff, S., Abraham, C.,
Bock, N., Kietzmann, S., Goedde, A., Toksoz, E., Droege, A., Krobitsch, S., Korn, B.,
Birchmeier, W., Lehrach, H., and Wanker, E. E. (2005) A human protein-protein interaction
network: a resource for annotating the proteome. Cell 122, 957-968.
9. Rual, J. F., Venkatesan, K., Hao, T., Hirozane-Kishikawa, T., Dricot, A., Li, N., Berriz, G.
F., Gibbons, F. D., Dreze, M., Ayivi-Guedehoussou, N., Klitgord, N., Simon, C., Boxem, M.,
Milstein, S., Rosenberg, J., Goldberg, D. S., Zhang, L. V., Wong, S. L., Franklin, G., Li, S.,
Albala, J. S., Lim, J., Fraughton, C., Llamosas, E., Cevik, S., Bex, C., Lamesch, P., Sikorski, R.
S., Vandenhaute, J., Zoghbi, H. Y., Smolyar, A., Bosak, S., Sequerra, R., Doucette-Stamm, L.,
Cusick, M. E., Hill, D. E., Roth, F. P., and Vidal, M. (2005) Towards a proteome-scale map of
the human protein-protein interaction network. Nature 437, 1173-1178.
10. Colland, F., Jacq, X., Trouplin, V., Mougin, C., Groizeleau, C., Hamburger, A., Meil, A.,
Wojcik, J., Legrain, P., and Gauthier, J. M. (2004) Functional proteomics mapping of a human
signaling pathway. Genome Res. 14, 1324-1332.
11. Tewari, M., Hu, P. J., Ahn, J. S., Ayivi-Guedehoussou, N., Vidalain, P. O., Li, S.,
Milstein, S., Armstrong, C. M., Boxem, M., Butler, M. D., Busiguina, S., Rual, J. F., Ibarrola, N.,
Chaklos, S. T., Bertin, N., Vaglio, P., Edgley, M. L., King, K. V., Albert, P. S., Vandenhaute, J.,
Pandey, A., Riddle, D. L., Ruvkun, G., and Vidal, M. (2004) Systematic interactome mapping
and genetic perturbation analysis of a C. elegans TGF-beta signaling network. Mol. Cell 13, 469-
12. Li, S., Armstrong, C. M., Bertin, N., Ge, H., Milstein, S., Boxem, M., Vidalain, P. O.,
Han, J. D., Chesneau, A., Hao, T., Goldberg, D. S., Li, N., Martinez, M., Rual, J. F., Lamesch, P.,
Xu, L., Tewari, M., Wong, S. L., Zhang, L. V., Berriz, G. F., Jacotot, L., Vaglio, P., Reboul, J.,
Hirozane-Kishikawa, T., Li, Q., Gabel, H. W., Elewa, A., Baumgartner, B., Rose, D. J., Yu, H.,
Bosak, S., Sequerra, R., Fraser, A., Mango, S. E., Saxton, W. M., Strome, S., Van Den Heuvel,
S., Piano, F., Vandenhaute, J., Sardet, C., Gerstein, M., Doucette-Stamm, L., Gunsalus, K. C.,
Harper, J. W., Cusick, M. E., Roth, F. P., Hill, D. E., and Vidal, M. (2004) A map of the
interactome network of the metazoan C. elegans. Science 303, 540-543.
13. Lim, J., Hao, T., Shaw, C., Patel, A. J., Szabo, G., Rual, J. F., Fisk, C. J., Li, N., Smolyar,
A., Hill, D. E., Barabasi, A. L., Vidal, M., and Zoghbi, H. Y. (2006) A protein-protein interaction
network for human inherited ataxias and disorders of Purkinje cell degeneration. Cell 125, 801-
14. Walhout, A. J., and Vidal, M. (2001) High-throughput yeast two-hybrid assays for large-
scale protein interaction mapping. Methods 24, 297-306.
15. Bader, G. D., Donaldson, I., Wolting, C., Ouellette, B. F., Pawson, T., and Hogue, C. W.
(2001) BIND--The Biomolecular Interaction Network Database. Nucleic Acids Res. 29, 242-245.
16. Zanzoni, A., Montecchi-Palazzi, L., Quondam, M., Ausiello, G., Helmer-Citterich, M.,
and Cesareni, G. (2002) MINT: a Molecular INTeraction database. FEBS Lett. 513, 135-140.
17. Mishra, G. R., Suresh, M., Kumaran, K., Kannabiran, N., Suresh, S., Bala, P.,
Shivakumar, K., Anuradha, N., Reddy, R., Raghavan, T. M., Menon, S., Hanumanthu, G., Gupta,
M., Upendran, S., Gupta, S., Mahesh, M., Jacob, B., Mathew, P., Chatterjee, P., Arun, K. S.,
Sharma, S., Chandrika, K. N., Deshpande, N., Palvankar, K., Raghavnath, R., Krishnakanth, R.,
Karathia, H., Rekha, B., Nayak, R., Vishnupriya, G., Kumar, H. G., Nagini, M., Kumar, G. S.,
Jose, R., Deepthi, P., Mohan, S. S., Gandhi, T. K., Harsha, H. C., Deshpande, K. S., Sarker, M.,
Prasad, T. S., and Pandey, A. (2006) Human protein reference database--2006 update. Nucleic
Acids Res. 34, D411-414.
18. Prieto, C., and De Las Rivas, J. (2006) APID: Agile Protein Interaction DataAnalyzer.
Nucleic Acids Res. 34, W298-302.
19. Lemonde, S., Turecki, G., Bakish, D., Du, L., Hrdina, P. D., Bown, C. D., Sequeira, A.,
Kushwaha, N., Morris, S. J., Basak, A., Ou, X. M., and Albert, P. R. (2003) Impaired repression
at a 5-hydroxytryptamine 1A receptor gene polymorphism associated with major depression and
suicide. J. Neurosci. 23, 8788-8799.
20. Jones, G., Moore, C., Hashemolhosseini, S., and Brenner, H. R. (1999) Constitutively
active MuSK is clustered in the absence of agrin and induces ectopic postsynaptic-like
membranes in skeletal muscle fibers. J. Neurosci. 19, 3376-3383.
21. Rual, J. F., Hirozane-Kishikawa, T., Hao, T., Bertin, N., Li, S., Dricot, A., Li, N.,
Rosenberg, J., Lamesch, P., Vidalain, P. O., Clingingsmith, T. R., Hartley, J. L., Esposito, D.,
Cheo, D., Moore, T., Simmons, B., Sequerra, R., Bosak, S., Doucette-Stamm, L., Le Peuch, C.,
Vandenhaute, J., Cusick, M. E., Albala, J. S., Hill, D. E., and Vidal, M. (2004) Human ORFeome
version 1.1: a platform for reverse proteomics. Genome Res. 14, 2128-2135.
22. Vidalain, P. O., Boxem, M., Ge, H., Li, S., and Vidal, M. (2004) Increasing specificity in
high-throughput yeast two-hybrid experiments. Methods 32, 363-370.
23. Venkatesan, K., Rual, J. F., Vazquez, A., Stelzl, U., Lemmens, I., Hirozane-Kishikawa,
T., Hao, T., Zenkner, M., Xin, X., Goh, K. I., Yildirim, M. A., Simonis, N., Heinzmann, K.,
Gebreab, F., Sahalie, J. M., Cevik, S., Simon, C., de Smet, A. S., Dann, E., Smolyar, A.,
Vinayagam, A., Yu, H., Szeto, D., Borick, H., Dricot, A., Klitgord, N., Murray, R. R., Lin, C.,
Lalowski, M., Timm, J., Rau, K., Boone, C., Braun, P., Cusick, M. E., Roth, F. P., Hill, D. E.,
Tavernier, J., Wanker, E. E., Barabasi, A. L., and Vidal, M. (2009) An empirical framework for
binary interactome mapping. Nat. Methods 6, 83-90.
24. psort http://mobyle.pasteur.fr/cgi-bin/MobylePortal/portal.py?form=psort.
25. Braun, P., Tasan, M., Dreze, M., Barrios-Rodiles, M., Lemmens, I., Yu, H., Sahalie, J. M.,
Murray, R. R., Roncari, L., de Smet, A. S., Venkatesan, K., Rual, J. F., Vandenhaute, J., Cusick,
M. E., Pawson, T., Hill, D. E., Tavernier, J., Wrana, J. L., Roth, F. P., and Vidal, M. (2009) An
experimentally derived confidence score for binary protein-protein interactions. Nat. Methods 6,
26. Boxem, M., Maliga, Z., Klitgord, N., Li, N., Lemmens, I., Mana, M., de Lichtervelde, L.,
Mul, J. D., van de Peut, D., Devos, M., Simonis, N., Yildirim, M. A., Cokol, M., Kao, H. L., de
Smet, A. S., Wang, H., Schlaitz, A. L., Hao, T., Milstein, S., Fan, C., Tipsword, M., Drew, K.,
Galli, M., Rhrissorrakrai, K., Drechsel, D., Koller, D., Roth, F. P., Iakoucheva, L. M., Dunker, A.
K., Bonneau, R., Gunsalus, K. C., Hill, D. E., Piano, F., Tavernier, J., van den Heuvel, S.,
Hyman, A. A., and Vidal, M. (2008) A protein domain-based interactome network for C. elegans
early embryogenesis. Cell 134, 534-545.
27. Ali, A., Hoeflich, K. P., and Woodgett, J. R. (2001) Glycogen synthase kinase-3:
properties, functions, and regulation. Chem. Rev. 101, 2527-2540.
28. Ohteki, T., Parsons, M., Zakarian, A., Jones, R. G., Nguyen, L. T., Woodgett, J. R., and
Ohashi, P. S. (2000) Negative regulation of T cell proliferation and interleukin 2 production by
the serine threonine kinase GSK-3. J. Exp. Med. 192, 99-104.
29. McManus, E. J., Sakamoto, K., Armit, L. J., Ronaldson, L., Shpiro, N., Marquez, R., and
Alessi, D. R. (2005) Role that phosphorylation of GSK3 plays in insulin and Wnt signalling
defined by knockin analysis. EMBO J. 24, 1571-1583.
30. Doble, B. W., Patel, S., Wood, G. A., Kockeritz, L. K., and Woodgett, J. R. (2007)
Functional redundancy of GSK-3alpha and GSK-3beta in Wnt/beta-catenin signaling shown by
using an allelic series of embryonic stem cell lines. Dev. Cell 12, 957-971.
31. He, X., Saint-Jeannet, J. P., Woodgett, J. R., Varmus, H. E., and Dawid, I. B. (1995)
Glycogen synthase kinase-3 and dorsoventral patterning in Xenopus embryos. Nature 374, 617-
32. Doble, B. W., and Woodgett, J. R. (2003) GSK-3: tricks of the trade for a multi-tasking
kinase. J. Cell Sci. 116, 1175-1186.
33. Mishra, R., Barthwal, M. K., Sondarva, G., Rana, B., Wong, L., Chatterjee, M.,
Woodgett, J. R., and Rana, A. (2007) Glycogen synthase kinase-3beta induces neuronal cell death
via direct phosphorylation of mixed lineage kinase 3. J. Biol. Chem. 282, 30393-30405.
34. Czesak, M., Lemonde, S., Peterson, E. A., Rogaeva, A., and Albert, P. R. (2006) Cell-
specific repressor or enhancer activities of Deaf-1 at a serotonin 1A receptor gene polymorphism.
J. Neurosci. 26, 1864-1871.
35. Kunick, C., Lauenroth, K., Leost, M., Meijer, L., and Lemcke, T. (2004) 1-
Azakenpaullone is a selective inhibitor of glycogen synthase kinase-3 beta. Bioorg. Med. Chem.
Lett. 14, 413-416.
36. Phiel, C. J., and Klein, P. S. (2001) Molecular targets of lithium action. Annu. Rev.
Pharmacol. Toxicol. 41, 789-813.
37. Huggenvik, J. I., Michelson, R. J., Collard, M. W., Ziemba, A. J., Gurley, P., and Mowen,
K. A. (1998) Characterization of a nuclear deformed epidermal autoregulatory factor-1 (DEAF-
1)-related (NUDR) transcriptional regulator protein. Mol. Endocrinol. 12, 1619-1639.
38. MacAulay, K., Doble, B. W., Patel, S., Hansotia, T., Sinclair, E. M., Drucker, D. J., Nagy,
A., and Woodgett, J. R. (2007) Glycogen synthase kinase 3alpha-specific regulation of murine
hepatic glycogen metabolism. Cell Metab. 6, 329-337.
39. Hoeflich, K. P., Luo, J., Rubie, E. A., Tsao, M. S., Jin, O., and Woodgett, J. R. (2000)
Requirement for glycogen synthase kinase-3beta in cell survival and NF-kappaB activation.
Nature 406, 86-90.
40. Phiel, C. J., Wilson, C. A., Lee, V. M., and Klein, P. S. (2003) GSK-3alpha regulates
production of Alzheimer's disease amyloid-beta peptides. Nature 423, 435-439.
41. Patel, S., Doble, B. W., MacAulay, K., Sinclair, E. M., Drucker, D. J., and Woodgett, J.
R. (2008) Tissue-specific role of glycogen synthase kinase 3beta in glucose homeostasis and
insulin action. Mol. Cell. Biol. 28, 6314-6328.
42. Liu, P., Cheng, H., Roberts, T. M., and Zhao, J. J. (2009) Targeting the phosphoinositide
3-kinase pathway in cancer. Nat. Rev. Drug Discov. 8, 627-644.
43. Samuels, Y., and Ericson, K. (2006) Oncogenic PI3K and its role in cancer. Curr. Opin.
Oncol. 18, 77-82.
44. Cuenda, A., and Rousseau, S. (2007) p38 MAP-kinases pathway regulation, function and
role in human diseases. Biochim. Biophys. Acta 1773, 1358-1375.
45. Rosner, M., Hanneder, M., Siegel, N., Valli, A., and Hengstschlager, M. (2008) The
tuberous sclerosis gene products hamartin and tuberin are multifunctional proteins with a wide
spectrum of interacting partners. Mutat. Res. 658, 234-246.
46. Patel, S., Doble, B., and Woodgett, J. R. (2004) Glycogen synthase kinase-3 in insulin and
Wnt signalling: a double-edged sword? Biochem. Soc. Trans. 32, 803-808.
47. Thyssen, G., Li, T. H., Lehmann, L., Zhuo, M., Sharma, M., and Sun, Z. (2006) LZTS2 is
a novel beta-catenin-interacting protein and regulates the nuclear export of beta-catenin. Mol.
Cell. Biol. 26, 8857-8867.
48. Chen, T., Li, M., Ding, Y., Zhang, L. S., Xi, Y., Pan, W. J., Tao, D. L., Wang, J. Y., and
Li, L. (2009) Identification of zinc-finger BED domain-containing 3 (Zbed3) as a novel Axin-
interacting protein that activates Wnt/beta-catenin signaling. J. Biol. Chem. 284, 6683-6689.
49. Krymskaya, V. P. (2003) Tumour suppressors hamartin and tuberin: intracellular
signalling. Cell. Signal. 15, 729-739.
50. Hadad, S. M., Fleming, S., and Thompson, A. M. (2008) Targeting AMPK: a new
therapeutic opportunity in breast cancer. Crit. Rev. Oncol. Hematol. 67, 1-7.
51. Inoki, K., Ouyang, H., Zhu, T., Lindvall, C., Wang, Y., Zhang, X., Yang, Q., Bennett, C.,
Harada, Y., Stankunas, K., Wang, C. Y., He, X., MacDougald, O. A., You, M., Williams, B. O.,
and Guan, K. L. (2006) TSC2 integrates Wnt and energy signals via a coordinated
phosphorylation by AMPK and GSK3 to regulate cell growth. Cell 126, 955-968.
52. Lee, J. H., Koh, H., Kim, M., Kim, Y., Lee, S. Y., Karess, R. E., Lee, S. H., Shong, M.,
Kim, J. M., Kim, J., and Chung, J. (2007) Energy-dependent regulation of cell structure by AMP-
activated protein kinase. Nature 447, 1017-1020.
53. Mirouse, V., Swick, L. L., Kazgan, N., St Johnston, D., and Brenman, J. E. (2007) LKB1
and AMPK maintain epithelial cell polarity under energetic stress. J. Cell Biol. 177, 387-392.
54. Zheng, B., and Cantley, L. C. (2007) Regulation of epithelial tight junction assembly and
disassembly by AMP-activated protein kinase. Proc. Natl. Acad. Sci. U. S. A. 104, 819-822.
55. Wang, C., Navab, R., Iakovlev, V., Leng, Y., Zhang, J., Tsao, M. S., Siminovitch, K.,
McCready, D. R., and Done, S. J. (2007) Abelson interactor protein-1 positively regulates breast
cancer cell proliferation, migration, and invasion. Mol. Cancer Res. 5, 1031-1039.
56. Yu, W., Sun, X., Clough, N., Cobos, E., Tao, Y., and Dai, Z. (2008) Abi1 gene silencing
by short hairpin RNA impairs Bcr-Abl-induced cell adhesion and migration in vitro and
leukemogenesis in vivo. Carcinogenesis 29, 1717-1724.
57. Kopp, P., Lammers, R., Aepfelbacher, M., Woehlke, G., Rudel, T., Machuy, N., Steffen,
W., and Linder, S. (2006) The kinesin KIF1C and microtubule plus ends regulate podosome
dynamics in macrophages. Mol. Biol. Cell 17, 2811-2823.
58. Grinberg-Rashi, H., Ofek, E., Perelman, M., Skarda, J., Yaron, P., Hajduch, M., Jacob-
Hirsch, J., Amariglio, N., Krupsky, M., Simansky, D. A., Ram, Z., Pfeffer, R., Galernter, I.,
Steinberg, D. M., Ben-Dov, I., Rechavi, G., and Izraeli, S. (2009) The expression of three genes
in primary non-small cell lung cancer is associated with metastatic spread to the brain. Clin.
Cancer Res. 15, 1755-1761.
59. Harada, M., Ishihara, Y., Itoh, K., and Yamanaka, R. (2007) Kinesin superfamily protein-
derived peptides with the ability to induce glioma-reactive cytotoxic T lymphocytes in human
leukocyte antigen-A24+ glioma patients. Oncol. Rep. 17, 629-636.
60. Hardie, D. G. (2005) New roles for the LKB1-->AMPK pathway. Curr. Opin. Cell Biol.
61. Lamia, K. A., Sachdeva, U. M., DiTacchio, L., Williams, E. C., Alvarez, J. G., Egan, D.
F., Vasquez, D. S., Juguilon, H., Panda, S., Shaw, R. J., Thompson, C. B., and Evans, R. M.
(2009) AMPK regulates the circadian clock by cryptochrome phosphorylation and degradation.
Science 326, 437-440.
62. Nakashima, A., Kawamoto, T., Honda, K. K., Ueshima, T., Noshiro, M., Iwata, T.,
Fujimoto, K., Kubo, H., Honma, S., Yorioka, N., Kohno, N., and Kato, Y. (2008) DEC1
modulates the circadian phase of clock gene expression. Mol. Cell. Biol. 28, 4080-4092.
63. Iizuka, K., and Horikawa, Y. (2008) Regulation of lipogenesis via BHLHB2/DEC1 and
ChREBP feedback looping. Biochem. Biophys. Res. Commun. 374, 95-100.
64. Cawley, N. X., Zhou, J., Hill, J. M., Abebe, D., Romboz, S., Yanik, T., Rodriguiz, R. M.,
Wetsel, W. C., and Loh, Y. P. (2004) The carboxypeptidase E knockout mouse exhibits
endocrinological and behavioral deficits. Endocrinology 145, 5807-5819.
65. Varlamov, O., Fricker, L. D., Furukawa, H., Steiner, D. F., Langley, S. H., and Leiter, E.
H. (1997) Beta-cell lines derived from transgenic Cpe(fat)/Cpe(fat) mice are defective in
carboxypeptidase E and proinsulin processing. Endocrinology 138, 4883-4892.
66. Chen, H., Jawahar, S., Qian, Y., Duong, Q., Chan, G., Parker, A., Meyer, J. M., Moore, K.
J., Chayen, S., Gross, D. J., Glaser, B., Permutt, M. A., and Fricker, L. D. (2001) Missense
polymorphism in the human carboxypeptidase E gene alters enzymatic activity. Hum. Mutat. 18,
67. Ma, L., Hanson, R. L., Que, L. N., Cali, A. M., Fu, M., Mack, J. L., Infante, A. M.,
Kobes, S., Bogardus, C., Shuldiner, A. R., and Baier, L. J. (2007) Variants in ARHGEF11, a
candidate gene for the linkage to type 2 diabetes on chromosome 1q, are nominally associated
with insulin resistance and type 2 diabetes in Pima Indians. Diabetes 56, 1454-1459.
68. Fu, M., Sabra, M. M., Damcott, C., Pollin, T. I., Ma, L., Ott, S., Shelton, J. C., Shi, X.,
Reinhart, L., O'Connell, J., Mitchell, B. D., Baier, L. J., and Shuldiner, A. R. (2007) Evidence
that Rho guanine nucleotide exchange factor 11 (ARHGEF11) on 1q21 is a type 2 diabetes
susceptibility gene in the Old Order Amish. Diabetes 56, 1363-1368.
69. Bottcher, Y., Schleinitz, D., Tonjes, A., Bluher, M., Stumvoll, M., and Kovacs, P. (2008)
R1467H variant in the rho guanine nucleotide exchange factor 11 (ARHGEF11) is associated
with impaired glucose tolerance and type 2 diabetes in German Caucasians. J. Hum. Genet. 53,
70. Ishikura, S., Koshkina, A., and Klip, A. (2008) Small G proteins in insulin action: Rab
and Rho families at the crossroads of signal transduction and GLUT4 vesicle traffic. Acta Physiol
(Oxf) 192, 61-74.
71. Deepa, S. S., and Dong, L. Q. (2009) APPL1: role in adiponectin signaling and beyond.
Am J Physiol Endocrinol Metab 296, E22-36.
72. Wang, C., Mao, X., Wang, L., Liu, M., Wetzel, M. D., Guan, K. L., Dong, L. Q., and Liu,
F. (2007) Adiponectin sensitizes insulin signaling by reducing p70 S6 kinase-mediated serine
phosphorylation of IRS-1. J. Biol. Chem. 282, 7991-7996.
73. Fruman, D. A., and Bismuth, G. (2009) Fine tuning the immune response with PI3K.
Immunol. Rev. 228, 253-272.
74. Weichhart, T., and Saemann, M. D. (2009) The multiple facets of mTOR in immunity.
Trends Immunol. 30, 218-226.
75. Karin, M., and Gallagher, E. (2009) TNFR signaling: ubiquitin-conjugated TRAFfic
signals control stop-and-go for MAPK signaling complexes. Immunol. Rev. 228, 225-240.
76. Jabara, H. H., Weng, Y., Sannikova, T., and Geha, R. S. (2009) TRAF2 and TRAF3
independently mediate Ig class switching driven by CD40. Int. Immunol. 21, 477-488.
77. Dupoux, A., Cartier, J., Cathelin, S., Filomenko, R., Solary, E., and Dubrez-Daloz, L.
(2009) cIAP1-dependent TRAF2 degradation regulates the differentiation of monocytes into
macrophages and their response to CD40 ligand. Blood 113, 175-185.
78. Karbowniczek, M., Cash, T., Cheung, M., Robertson, G. P., Astrinidis, A., and Henske, E.
P. (2004) Regulation of B-Raf kinase activity by tuberin and Rheb is mammalian target of
rapamycin (mTOR)-independent. J. Biol. Chem. 279, 29930-29937.
79. Vinuesa, C. G., Cook, M. C., Angelucci, C., Athanasopoulos, V., Rui, L., Hill, K. M., Yu,
D., Domaschenz, H., Whittle, B., Lambe, T., Roberts, I. S., Copley, R. R., Bell, J. I., Cornall, R.
J., and Goodnow, C. C. (2005) A RING-type ubiquitin ligase family member required to repress
follicular helper T cells and autoimmunity. Nature 435, 452-458.
80. Yu, D., Tan, A. H., Hu, X., Athanasopoulos, V., Simpson, N., Silva, D. G., Hutloff, A.,
Giles, K. M., Leedman, P. J., Lam, K. P., Goodnow, C. C., and Vinuesa, C. G. (2007) Roquin
represses autoimmunity by limiting inducible T-cell co-stimulator messenger RNA. Nature 450,
81. Beaulieu, J. M., Gainetdinov, R. R., and Caron, M. G. (2008) Akt/GSK3 Signaling in the
Action of Psychotropic Drugs. Annu. Rev. Pharmacol. Toxicol.
82. Jackson, M., Song, W., Liu, M. Y., Jin, L., Dykes-Hoberg, M., Lin, C. I., Bowers, W. J.,
Federoff, H. J., Sternweis, P. C., and Rothstein, J. D. (2001) Modulation of the neuronal
glutamate transporter EAAT4 by two interacting proteins. Nature 410, 89-93.
83. Perrot, V., Vazquez-Prado, J., and Gutkind, J. S. (2002) Plexin B regulates Rho through
the guanine nucleotide exchange factors leukemia-associated Rho GEF (LARG) and PDZ-
RhoGEF. J. Biol. Chem. 277, 43115-43120.
84. Simpson, M. A., Cross, H., Proukakis, C., Pryde, A., Hershberger, R., Chatonnet, A.,
Patton, M. A., and Crosby, A. H. (2003) Maspardin is mutated in mast syndrome, a complicated
form of hereditary spastic paraplegia associated with dementia. Am. J. Hum. Genet. 73, 1147-
85. Karlstrom, H., Brooks, W. S., Kwok, J. B., Broe, G. A., Kril, J. J., McCann, H., Halliday,
G. M., and Schofield, P. R. (2008) Variable phenotype of Alzheimer's disease with spastic
paraparesis. J. Neurochem. 104, 573-583.
86. Drevets, W. C., Thase, M. E., Moses-Kolko, E. L., Price, J., Frank, E., Kupfer, D. J., and
Mathis, C. (2007) Serotonin-1A receptor imaging in recurrent depression: replication and
literature review. Nucl. Med. Biol. 34, 865-877.
87. Celada, P., Puig, M., Amargos-Bosch, M., Adell, A., and Artigas, F. (2004) The
therapeutic role of 5-HT1A and 5-HT2A receptors in depression. J. Psychiatry Neurosci. 29, 252-
88. Sharp, T., Boothman, L., Raley, J., and Queree, P. (2007) Important messages in the
'post': recent discoveries in 5-HT neurone feedback control. Trends Pharmacol. Sci. 28, 629-636.
89. Savitz, J., Lucki, I., and Drevets, W. C. (2009) 5-HT(1A) receptor function in major
depressive disorder. Prog. Neurobiol. 88, 17-31.
90. Drevets, W. C., Frank, E., Price, J. C., Kupfer, D. J., Holt, D., Greer, P. J., Huang, Y.,
Gautier, C., and Mathis, C. (1999) PET imaging of serotonin 1A receptor binding in depression.
Biol. Psychiatry 46, 1375-1387.
91. Moses-Kolko, E. L., Wisner, K. L., Price, J. C., Berga, S. L., Drevets, W. C., Hanusa, B.
H., Loucks, T. L., and Meltzer, C. C. (2008) Serotonin 1A receptor reductions in postpartum
depression: a positron emission tomography study. Fertil. Steril. 89, 685-692.
92. Bain, E. E., Nugent, A. C., Carson, R. E., Luckenbaugh, D., Lang, L., Eckekman, W. C.,
Neumeister, A., Bonne, O., Williams, J., Gordon, J., Charney, D. S., and Drevets, W. C. (2004)
Decreased 5-HT1A receptor binding in bipolar depression. Biol. Psychiatry 55, 178S.
93. Szewczyk, B., Albert, P. R., Burns, A. M., Czesak, M., Overholser, J. C., Jurjus, G. J.,
Meltzer, H. Y., Konick, L. C., Dieter, L., Herbst, N., May, W., Rajkowska, G., Stockmeier, C. A.,
and Austin, M. C. (2008) Gender-specific decrease in NUDR and 5-HT1A receptor proteins in
the prefrontal cortex of subjects with major depressive disorder. Int. J. Neuropsychopharmacol.,
94. Sargent, P. A., Kjaer, K. H., Bench, C. J., Rabiner, E. A., Messa, C., Meyer, J., Gunn, R.
N., Grasby, P. M., and Cowen, P. J. (2000) Brain serotonin1A receptor binding measured by
positron emission tomography with [11C]WAY-100635: effects of depression and antidepressant
treatment. Arch. Gen. Psychiatry 57, 174-180.
95. Meltzer, C. C., Price, J. C., Mathis, C. A., Butters, M. A., Ziolko, S. K., Moses-Kolko, E.,
Mazumdar, S., Mulsant, B. H., Houck, P. R., Lopresti, B. J., Weissfeld, L. A., and Reynolds, C.
F. (2004) Serotonin 1A receptor binding and treatment response in late-life depression.
Neuropsychopharmacology 29, 2258-2265.
96. Beaulieu, J. M., Marion, S., Rodriguiz, R. M., Medvedev, I. O., Sotnikova, T. D., Ghisi,
V., Wetsel, W. C., Lefkowitz, R. J., Gainetdinov, R. R., and Caron, M. G. (2008) A beta-arrestin
2 signaling complex mediates lithium action on behavior. Cell 132, 125-136.
97. Beaulieu, J. M., Zhang, X., Rodriguiz, R. M., Sotnikova, T. D., Cools, M. J., Wetsel, W.
C., Gainetdinov, R. R., and Caron, M. G. (2008) Role of GSK3 beta in behavioral abnormalities
induced by serotonin deficiency. Proc. Natl. Acad. Sci. U. S. A. 105, 1333-1338.
98. Kang, U. G., Seo, M. S., Roh, M. S., Kim, Y., Yoon, S. C., and Kim, Y. S. (2004) The
effects of clozapine on the GSK-3-mediated signaling pathway. FEBS Lett. 560, 115-119.
99. Karege, F., Perroud, N., Burkhardt, S., Schwald, M., Ballmann, E., La Harpe, R., and
Malafosse, A. (2007) Alteration in kinase activity but not in protein levels of protein kinase B
and glycogen synthase kinase-3beta in ventral prefrontal cortex of depressed suicide victims.
Biol. Psychiatry 61, 240-245.
100. Carlson, P. J., Bain, E., Tinsley, R., Luckenbaugh, D., Manji, H. K., and Drevets, W. C.
(2007) Serotonin-1A receptor binding in bipolar depression before and after mood stabilizer
treatment. Biol. Psychiatry 61, 57S.
Description of the PI3K-mTOR pathway A- Functional relationships between the proteins of
PI3K-mTOR pathway and major outputs of the pathway activation. Modules are color-coded:
orange, green, pink and blue correspond to the receptor, the PI3K, the mTOR and the related
proteins modules respectively. Functional paralogs are depicted as joined rectangles. PI:
phosphatidylinositol; PIP: phosphatidylinositide phosphate. B- Current designations of the PI3K-
mTOR pathway components are presented in parallel to their official symbols and to their gene
ID. Genes indicated in italic could not be included in our yeast-two hybrid screen but were
integrated in the literature-completed interactome.
Yeast-two hybrid interaction network for the PI3K-mTOR pathway Y2H baits are depicted
are triangles and the discovered interactors as circles. Baits for which no interaction was found
are not represented. Colour of the links is indicative of the Y2H method that led to its discovery.
Thickness of the links is indicative of the number of times an interaction was found in the Y2H
cDNA or high throughput (HTP) screen. The molecular function of each protein is described by a
Functional annotation of the literature-completed interactome of the PI3K-mTOR pathway
PI3K-mTOR pathway components are depicted are triangles and interactors as circles. Pink links
represent interactions found in this study and grey links, manually-curated interactions from
literature. The molecular function of each protein is described by a colour code and interactors
are grouped according to their sub-cellular functions.
Connections of the literature-completed interactome of the PI3K-mTOR pathway with
other signaling pathways PI3K-mTOR pathway components are depicted are triangles and
interactors as circles. Pink links represent interactions found in this study and grey links,
manually-curated interactions from literature. The molecular function of each protein is described
by a colour code and interactors are grouped according to their belonging to specific signaling
Interactions for GSK3A and GSK3B variant constructs A- The different constructs used for
Y2H and co-AP experiments are presented. The black lines represent the homolog parts of the
proteins whereas the divergent parts are in grey. Amino acids S21 and S9 are inhibitory sites and
K148 and K85 are essential residues for the catalytic sites in GSK3A and GSK3B respectively.
Mutated residues are underlined. B- The table recapitulates results obtained for Y2H experiments
with the different GSK3A and GSK3B constructs. The experiments were performed twice, except
for GSK3A-SA-KA and GSK3B-SA-KA constructs that were tested once. A positive/negative
result reflects a discrepancy between both experiments, except for GSK3AΔ-C14orf129
interaction that was twice at the positive threshold. C- Co-AP results for a set of interactors are
presented. Experiments were done once and a positive/negative result reflects an ambiguous
profile for Myc detection in eluates.
DEAF1 in vitro phosphorylation by GSK3A and GSK3B kinases A- DEAF1-FL, DEAF1-1,
DEAF1-2, DEAF1-3 constructs used for kinase assay are represented in parallel to the cDNA
clone that was isolated from Y2H screen. Numbers in italic indicate the first and last amino acids
of the constructs in comparison to DEAF1-FL. Conserved residues between human and mouse
that are potential sites of phosphorylation by GSK3A and GSK3B are quoted on DEAF1-FL
construct. B- Quantification of purified wild-type and kinase-dead (KA) GST-GSK3A and GST-
GSK3B, as well as M2-MLK3, by western-blot with anti-GST and -M2 antibodies respectively.
C- Upper parts: 32P radioactive signal from kinase assay reveals MLK3 (positive control),
DEAF1-FL, DEAF1-1, DEAF1-2, DEAF1-3 proteins phosphorylation by wild type GSK3A and
GSK3B kinases. Lower parts: quantitative fluorescent western blot for Myc on the kinase assay
membranes revealed the respective amount of Myc-DEAF1 proteins. D- Adjusted quantification
of DEAF1 constructs phosphorylation by GSK3A and GSK3B was calculated as the ratio (32P
radioactive signal/Myc signal quantification) for each DEAF1 construct. a.u.: arbitrary units.
Activation of 5HT1A promoter by DEAF1 and GSK3 inhibitors A,B- Results of luciferase
assays on HEK293T cells (A) and RN46A cells (B) show the influence of DEAF1 on 5HT1A
promoter activity, alone or in addition to GSK3 inhibitors, i.e. lithium chloride and
azakenpaullone, in comparison to control conditions (Myc vector, sodium chloride and
dimethylsulfoxide respectively). C, D- Results of luciferase assays on HEK293T cells (C) and
RN46A cells (D) show the influence of GSK3 inhibitors on 5HT1A promoter activity in
comparison to control conditions. Independent experiments were performed at least 6 times.
Luciferase level was adjusted to transfection conditions (see material and methods). Histograms
show means +/- SEM. Statistical analysis was performed using Student t-test. * = p<0.05; ** =
p<0.025; *** = p<0.01; **** = p<0.005; ***** = p<0.001.
Page 50 Download full-text