1

TITLE PAGE

Full title: The tyrosine kinase receptor RET interacts in vivo with AIP to alter survivin availability.

Authors’ names and institutions:

Manuela Vargiolu

Unità di Genetica Medica, Policlinico Universitario S. Orsola-Malpighi, Bologna, Italy

Daniela Fusco

Unità di Genetica Medica, Policlinico Universitario S. Orsola-Malpighi, Bologna, Italy

Ivana Kurelac

Unità di Genetica Medica, Policlinico Universitario S. Orsola-Malpighi, Bologna, Italy

Dietmar Dirnberger

Bio3/Bioinformatics and Molecular Genetics (Faculty of Biology), University of Freiburg, Germany.

Ralf Baumeister

Bio3/Bioinformatics and Molecular Genetics (Faculty of Biology),Center for Biochemistry and

Molecular Cell Research (Faculty of Medicine), Center for Systems Biology (ZBSA) and FRIAS,

School of Life Sciences (LIFENET), Albert-Ludwigs-University Freiburg, Schänzlestr. 1, D79104

Freiburg i. Brsg., Germany

Isabella Morra

Department of Histopathology, Ospedale Infantile Regina Margherita, Torino, Italy

Antonio Melcarne

Department of Neurosurgery, A.S.O. CTO-CRF-M.Adelaide, Torino, Italy

Roberto Rimondini

Dept. of Pharmacology, University of Bologna, Via Irnerio 48 Bologna, Italy

J Clin Endocrin Metab. First published ahead of print April 14, 2009 as doi:10.1210/jc.2008-1980

Copyright (C) 2009 by The Endocrine Society

2

Giovanni Romeo

Unità di Genetica Medica, Policlinico Universitario S. Orsola-Malpighi, Bologna, Italy

Elena Bonora

Unità di Genetica Medica, Policlinico Universitario S. Orsola-Malpighi, Bologna, Italy

Abbreviated title (max 40 chrs): AIP interacts with RET

Word count excluding abstract, figure captions, ref: 3599

Name and address of person to whom reprint requests should be addressed: Giovanni Romeo

<romeo@eurogene.org

Any grants or fellowships supporting the writing of the paper: European Specific Targeted

Research Projects project HERMIONE, University of Turin, University of Bologna, the EC 6th

Framework Network of Excellence LIFESPAN and a program of the German Federal Ministry in

Research and Education.

Author Disclosure Summary: M.V., D.F., I.K., D.D., R.B., I.M., A.M., R.R., G.R., E.B. have nothing

to declare. G.R. received grant support (2006-2009) from EU.

Precis: The identification of the AIP-RET complex represents a starting point to study key cellular

processes involved in RET induced apoptosis.

3

ABSTRACT (230 words)

Context: RET is a tyrosine kinase transmembrane receptor expressed in two main alternative isoforms:

RET9 and RET51. RET transduces a positive signal leading to survival, differentiation or migration in

the presence of its ligand GDNF, whilst in its absence a proapoptotic fragment which initiates a

negative signalling for apoptosis is generated. The signal transduction mechanisms leading to apoptosis

are still unclear.

Objective: In order to shed light on the mechanisms of RET induced apoptosis we searched for novel

interactors of RET51.

Design: The “split ubiquitin yeast two hybrid system” was used with RET51 as bait against a human

brain expression library.

Results: We identified AIP, a co-chaperone recently found mutated in pituitary adenoma patients, as a

novel interactor of RET. We showed that RET interacts specifically with AIP both in mammalian cell

lines and in vivo in the pituitary gland, regardless of the presence of pituitary adenoma specific

mutations. AIP and RET genes were sequenced in 28 pituitary adenoma but no relevant mutation were

found. In addition, we identified the pro-apoptotic domain of RET as responsible for the interaction

with AIP. Finally, we demonstrated that the AIP-RET interaction does not require RET kinase activity

or kinase dependent signal transduction and that it prevents the formation of the AIP-survivin complex.

Conclusions: The identification of the AIP-RET complex represents a starting point to study key

cellular processes involved in RET induced apoptosis.

4

INTRODUCTION

The RET proto-oncogene is the tyrosine kinase receptor for the glial cell line-derived neurotrophic

factor (GDNF) family of ligands. Binding occurs in conjunction with glycosylphosphatidyl inositol-

membrane anchored co-receptors designated GDNF-family receptor-α (GFRα)(1). RET is expressed in

two main isoforms: the short and the long isoforms, namely RET9 and RET51, which differs at the C-

terminal tail(2). Gain- and loss-of-function mutations of RET have been associated respectively with

neoplasia (Multiple Endocrine Neoplasia Type 2A–MEN2A-and 2B–MEN2B-and Familial Medullary

Thyroid Carcinoma–FMTC-) and with Hirschsprung disease, a neurodevelopmental disorder.

RET has a dual role according to the presence or absence of GDNF. In the presence of GDNF, RET

promotes survival, growth and migration of cells such as neurons and epithelial renal cells. The survival

signal transduced by RET is required for maturation of several cell lineages of the peripheral nervous

system, kidney morphogenesis, and spermatogenesis(3). Conversely, in the absence of its ligand RET is

able to induce programmed cell death by releasing a proapoptotic fragment which initiates a negative

signal for apoptosis, as shown in Neuro2A cells and in transient transfected HEK293T(4) and GH4C1

cell lines(5). The precise mechanisms of RET induced apoptosis remain nevertheless unclear.

In order to shed light on the proteins involved in the pathway induced by RET, we used a modified

yeast two hybrid method named the“split ubiquitin system”(6). This method led to the identification of

a novel interactor of RET51 among human brain proteins. We here identify a novel binding partner of

RET, the aryl hydrocarbon receptor interacting protein (AIP), also known as ARA9 or XAP2(7, 8), a

tumor-suppressor protein recently found mutated in pituitary adenoma. We show that RET-AIP

interaction is maintained both in cell lines and in the pituitary gland in vivo, regardless of the presence

of pituitary adenoma specific mutations, possibly altering the cellular apoptotic potential.

5

MATERIALS AND METHODS

Split Ubiquitin Yeast Two Hybrid

The bait construct was based on the previously described vector pPCUP1-ubc9-CRU (6), which is a

yeast–E.coli shuttle vector that carries a CRU locus, a copper-dependent promoter (PCUP1) for bait

expression, and a HIS3 marker. The human coding sequence corresponding to RET51 intracellular

domain excluding the transmembrane segment was amplified by PCR from a human whole brain cDNA

preparation (Clontech) using primers RET51-FW(5’-

AATTGTCGACATGCACTGCTACCACAAGTTTGCCC-3’) and RET51-RV(5’-

AGCGGCCGCACTATCAAACGTGTCCATTA-3’). The RET51 PCR product was used to replace the

Ubc9 insert in pPCUP1-ubc9-CRU, resulting in pPCUP1-Fz1-CRU. The human fetal brain NubI—split

ubiquitin cDNA library was kindly provided by GPC Biotech AG. The cDNA library screening and the

selection of positive clones was performed as previously described (6).

Cell cultures and transfections

HEK293, Neuro2A and SH-SY5Y cells were grown at 37°C with 5% CO2 in DMEM (Celbio)

supplemented with 10% fetal bovine serum (Euroclone), 2 mM L-glutamine (Euroclone), 100 units/ml

penicillin and 100 µg/ml streptomycin (Euroclone). Cells were plated on 60 mm Petri dishes and

transfection was performed with Lipofectamine Reagent (Invitrogen) following the manufacturer's

instructions.

cDNA construct preparation

Total RNA extracted from the SH-SY5Y neuroblastoma cell line using Trizol Reagent (Invitrogen)

was retrotranscribed using SuperScript™III Reverse Transcriptase (Invitrogen). The human AIP coding

sequence was amplified from cDNA by PCR using primers EcoRI_AIP-F(5’-

CCCGAATTCGCCGAAGCAAGTCCG-3’) and XhoI_AIP-R (5’-

GGCTCGAGCATGGGAGAAGATCCC-3’) and cloned into pcDNA3.1myc/His expression vector

6

(Invitrogen). The plasmids pJ7Ω-RET51 (RET51-FL) and pJ7Ω-RET9 (RET9-FL) were kindly

provided by P. Mehlen, (Apoptosis Cancer and Development Laboratory, Lyon, France;(4)) Single base

mutations were introduced into wild-type AIP and RET51 clones using the QuikChange site-directed

mutagenesis kit (Stratagene).

The construct containing tag-free AIP was produced by subcloning the AIP coding sequence from

pcDNA3.1myc/His expression vector (Invitrogen) to pcDNA3.1(+) (Invitrogen) using the primers

EcoRI_AIP-F(5’-CCCGAATTCGCCGAAGCAAGTCCG-3’) and XhoI_ext_AIP(5’-

GGGTGACCTCGAGTCAATGGGAGAAGATCCC-3’).

The deleted constructs RET-TK (aa 1-999) and RET-IUXTA (aa 1-725) were generated by PCR

amplification using primers RET_HindIII_F(5’-GGGATATCCCATGGCGAAGGCGACG-3’),

RET_IUXTA_XbaI_R(5’-GGTCTAGAAGAACCAAGTTCTTCCGAGG-3’), RET_TK_XbaI_R(5’-

GGTCTAGAGCCGCAAACACCGGCCTTTTGTCCG-3’) and cloned into pcDNA3.1/V5-His©

TOPO® TA Expression Vector (Invitrogen). RET-PRO (708-1017) was generated using primers

RET_HindIII_F(5’-GGGATATCCCATGGCGAAGGCGACG-3’) and NotI_PRORET_R(5’-

GCGGCCGCGTCCAAGTAGTCTCTCCT-3’) and cloned into pcDNA3.1myc/His expression vector

(Invitrogen).

All plasmid sequences were verified by direct sequencing using Big Dye v3.1 kit according to the

manufacturer’s instruction and loaded onto a 3730 automated sequencer (Applied Biosystems).

7

Immunoprecipitation assays and Western blot analysis

Cells were lysed in 100 µl of IP Buffer 1X (Sigma-Aldrich) containing protease Inhibitors (Roche

Diagnostics). Rat pituitary glands were homogenized in ice-cold IP Buffer 1X containing protease

Inhibitors (Roche Diagnostics). Cell lysates were incubated with 1 µg of primary antibody. After 4 hrs

20 µl of Protein G beads (Sigma-Aldrich) equilibrated in IP buffer were added and incubated overnight.

Bead pellets were washed 3 times with 200 µl of IP Buffer and resuspended in 40 µl of Laemmli buffer

(8% SDS; 20% glycerol; 125mM Tris HCl, pH 6.8; 0.0025% bromophenol blue; 10% 2-

mercaptoethanol). Immunoprecipitates and total cell lysates were separated by SDS-PAGE and

transferred onto nitrocellulose membrane. Immunodetection and immunoprecipitation were performed

using primary antibodies specific for RET (Ret(C-19) sc-167; Ret(C-20) sc-1290; Ret(H-300) sc-

13104), AIP (XAP2(35-2) sc-59730; XAP2(N-20) sc-27445) from Santa Cruz Biotechnology, for

survivin (GTX10584) from GeneTex and for the myc epitope (R950-25) from Invitrogen. Secondary

antibodies were HRP-conjugate anti-mouse or anti-rabbit (Sigma-Aldrich) HRP-conjugate anti-goat

(DAKO) and AP-conjugate anti-mouse (Invitrogen). Development was performed using ECL system

(Millipore).

Patient material and mutation screening

A total of 28 bioptic samples of sporadic pituitary adenoma were obtained from the Department of

Histopathology of Turin and kept anonymous. DNA was extracted from frozen or paraffin-embedded

pathological pituitary tissue using the NucleoSpin Tissue Kit (Mackerey-Nagel).

The whole AIP and RET coding sequences were analyzed for mutations by direct sequencing using Big

Dye v3.1 and 3730 automated sequencer (Applied Biosystems). The PCR conditions are available

under request. In silico analysis was performed using TargetScan 4.2(9), Mireval(10) and miRBase

v5(11) for miRNA binding site predictions and mfold 3.2 (12)for RNA folding analysis.

8

RESULTS

Identification of AIP as a binding partner of the long isoform of RET–In order to better characterise the

signaling pathway downstream of RET51, we performed a split ubiquitin yeast two-hybrid screening

using the cytoplasmic domain of RET51 as a bait against a human fetal brain cDNA library(6). We

employed the split-ubiquitin system based on the R-URA3p reporter which allowed positive selection

for uracil auxotrophy and negative selection in the presence of the otherwise toxic URA3p-specific

antimetabolite 5-fluoroorotic acid (5-FOA).

We used the full length cDNA of human RET51 to generate the pPCUP1-RET51-CRU bait expression

vector in which full-length RET51 cDNA was expressed as a translational fusion protein to URA3.

However, the transformants with pPCUP1-RET51-CRU did not grow on the appropriate medium. Thus,

the cytoplasmatic portion of RET51 was used as bait and we screened for protein interactors of this

fragment using a poly(T)-primed human fetal brain cDNA-library.

After eliminating false-positive interaction candidates, we isolated 10 URA-auxotrophic and 5-FOA

resistant clones. The 10 positive clones were sequenced, and nucleotide sequence database searches

were performed using BLAST (National Center for Biotechnology Information—NCBI). One of these

clones contained in the prey plasmid the cDNA encoding the entire protein of Aryl Hydrocarbon

Receptor Interacting Protein (NM_003977).

AIP interacts both with RET9 and RET51 in mammalian cell lines–In order to validate the AIP-RET51

interaction observed in yeast we performed a co-immunoprecipitation assay in mammalian cell lines.

Full-length RET51 (RET51-FL)(Fig.1) and AIP myc-tagged were expressed in 293 human embryonic

kidney (HEK) cells (Fig.2A). Cells transfected either with a mock or AIP expressing vector and

RET51-FL were lysed and immunoprecipitated with an antibody specific for the long isoform of RET

(Fig.2A). When the total cell lysates and the corresponding immunoprecipitates were analysed by SDS-

PAGE, the presence of a 39 KDa protein (the expected size of myc-tagged AIP) was detected both in

the total cell lysate and in the lysate immunoprecipitated with anti-RET51 antibody (Fig.2A). The 39

KDa protein was not detected when the mock plasmid was transfected (Fig.2A, third lane) thus

9

indicating that anti-myc antibody specifically detected AIP. Also, when lysates obtained from cells

transfected with the mock plasmid were immunoprecipitated with the anti-RET51 antibody, the 39 KDa

band was absent (Fig.2A, fourth lane). This demonstrated that AIP and RET51 co-immuprecipitated in

HEK293, indicating that AIP was able to bind RET51 in human cells.

To understand whether the AIP-RET51 interaction was also present in cells expressing RET

endogenously, the same experiment was performed in two neuroblastoma cell lines, human SH-SY5Y

and murine Neuro2A (Fig.2C,D). Immunoblotting confirmed the expression of AIP as well as RET in

both neuroblastoma cell lines (Fig. 2C). Neuro2A cells transfected either with a mock plasmid or with

the vector expressing myc-tagged AIP were lysated and immunoprecipitated with the antibody specific

for RET51 (Fig. 2D). As expected, when the cell lysates and the corresponding immunoprecipitates

were analysed by immunoblotting with an anti-myc antibody, the presence of AIP was detected both in

the total cell lysate and in the lysate immunoprecipitated with the anti-RET51 antibody (Fig. 2D, first

and second lane), whereas it was not detected when the mock plasmid was transfected (Fig. 2D, third

and fourth lane). In addition, when the cell lysate of non-transfected Neuro2A were immunoprecipitated

with the anti-RET51 antibody, the presence of endogenous AIP was detected (Fig. 2D, fifth and sixth

lane). Indeed, endogenous AIP was able to immunoprecipitate RET demonstrating that the AIP-RET

interaction is not dependent upon transfection of these proteins. The same results were obtained in

SHSY-5Y (data not shown).

To address the specificity of AIP binding for the long isoform of RET, the same co-

immunoprecipitation assay was performed in HEK293, transfected with full-length RET9 (RET9-FL)

(Fig.1) and myc-tagged AIP. Cells transfected either with a mock or AIP-expressing vector and RET9-

FL were lysated and immunoprecipitated with an antibody specific for RET9 (Fig.2B). Also in this

case, the presence of AIP was detected both in the total cell lysate and in the lysate immunoprecipitated

with the anti-RET9 antibody (Fig.2B, first and second lane), whereas it was not detected when the

mock plasmid was transfected (Fig.2B, third and fourth lane). This indicated that AIP was able to bind

to both RET isoforms in a region likely common to RET9 and RET51.

10

AIP interacts with the proapoptotic domain of RET–To determine the domain of RET involved in

binding to AIP we generated various constructs containing different functional domains of RET. RET-

IUXTA (aa 1-725) included the juxtamembrane domain of RET, known to have a role in regulating the

kinase activity of RET(13), RET-PRO (aa 707-1017) contained the proapoptotic domain of RET tagged

with myc and RET-TK (aa 1-999) encompassed both the juxtamembrane and the tyrosine kinase

domain of RET (Fig.1).

Cells transfected with either a mock or myc-tagged AIP expressing vector and either RET-IUXTA or

RET-TK were lysed and immunoprecipitated with an antibody specific for the N-terminal portion of

RET. AIP was present both in the cell lysate and in the lysate immunoprecipitated with the anti-RET

antibody when RET-TK was expressed, whereas it was not detected in the immunoprecipitates when

the RET-IUXTA plasmid was transfected (Fig.2E). Accordingly, in a reverse experiment using anti-

myc antibody for immunoprecipitation, we could detect RET by immunoblotting only in the

immunoprecipitates of cells expressing RET-TK (Fig.2F).

To further refine the mapping of the AIP-RET interaction, the proapoptotic domain of RET was tested

for its ability to bind AIP. HEK293 transfected with a mock or AIP-expressing vector and either a mock

or RET-PRO myc-tagged vector were lysated and immunoprecipitated with anti-AIP antibody. The

presence of RET-PRO was detected in the immunoprecipitates when AIP was also expressed, whereas

no protein could be revealed in the immunoprecipitates when each of the two mock plasmids were

expressed (Fig.2G). These data demonstrated that the portion of RET interacting with AIP spans from

amino acid 707 to 999, a region common to both isoforms.

AIP and RET interact with each other in vivo in the pituitary gland–The above results indicated that the

AIP-RET interaction was present in cells of different origin. In order to investigate whether this

interaction occurs in vivo, we analysed the rat pituitary, given that RET, GDNF and AIP are expressed

in somatotrophs and AIP is involved in pituitary tumor pathogenesis. To verify the endogenous

interaction between RET and AIP we performed co-immunoprecipitation experiments in lysates

prepared from rat pituitary gland tissue. Analysis of the total lysate by immunoblotting confirmed the

11

expression of both proteins (Fig.3A). Using a specific antibody against AIP, both AIP and RET could

be detected in the immunoprecipitates (Fig. 3B). No unspecific binding of the proteins occurred in a

control experiment using nonspecific IgG (Fig. 3, third lane). These results clearly proved the presence

of a AIP-RET complex in the pituitary gland.

RET phosphorylation status does not affect the AIP-RET interaction–In our experiments we transiently

over-expressed RET, a condition known to result in high levels of RET autophosphorylation in the

absence of ligand stimulation(14, 15). To investigate whether the interaction of RET with AIP relied on

a kinase-dependent signal transduction, we checked the ability of RET to retain the interaction with AIP

when its kinase activity was abolished. For this purpose cells expressing AIP and a mutated form of

RET containing a point mutation of the lysine in position 758 that renders RET kinase inactive (RET51-

DK, K758R)(14) were studied. The presence of AIP was detected in the immunoprecipitates of cells

expressing either RET51-FL or RET51-DK, suggesting that the AIP-RET interaction is not dependent

on RET phosphorylation status (Fig.4A).

RET51 prevents AIP binding to the anti-apoptotic protein survivin – In a previous study transient

transfection of the full-length RET in human embryonic kidney cells was shown to induce apoptosis

(4). AIP also has a role in regulating the cellular apoptotic threshold, directly binding and regulating the

stability of survivin, a member of the IAP (inhibitor of apoptosis) family(16). To investigate the

connection between these two processes, we tested the ability of transiently transfected RET in HEK to

affect the formation of the AIP-survivin complex. Cells over-expressing RET51 were lysed and the

ability of AIP to retain the interaction with survivin was assessed by immunoprecipitating with anti-

survivin antibody (Fig.4B). The presence of AIP in the immunoprecipitates was detected only in

absence of RET. In contrast, the presence of RET abolished the ability of AIP to co-immunoprecipitate

with survivin, demonstrating that RET was sufficient to prevent the formation of the AIP-survivin

complex (Fig.4B).

Pathogenic mutations of RET and AIP do not impair the AIP-RET interaction–Germline mutations of

AIP were recently described in sporadic and familial pituitary adenoma (17). The six more frequent

12

missense mutations found in pituitary adenoma patients were selected, namely R16H, K241E, R271W,

E293G, A299V, R304Q, and introduced in our construct containing AIP (Fig.5A). The resulting

constructs were expressed in HEK293 together with either RET51-FL or RET-9. Their ability to modify

the AIP-RET interaction was tested by immunoprecipitating the total cell lysate with an anti-RET

antibody and immonoblotting for AIP. The presence of all the six mutated AIP proteins was detected in

the lysate immunoprecipitated with an antibody specific for either RET51 (Fig.5B) or RET9 (data not

shown). These data clearly showed that none of these six missense mutations impairs the interaction of

AIP with RET51 or with RET9.To further prove that none of these AIP mutations impaired the

interaction with RET, the ability of the proapoptotic fragment of RET to bind the mutated forms of AIP

was tested. Cells expressing myc-tagged RET-PRO and each of the six mutated forms of AIP were

lysed and immunoprecipitated with the antibody specific for AIP. As expected, immunoblotting for the

myc tag of RET-PRO showed the presence of RET-PRO when immunoprecipitating the lysates of cells

transfected with each mutated form of AIP (Fig.5C). We therefore conclude that AIP pituitary adenoma

predisposing mutations do not impair AIP-RET interaction.

To further address the involvement of the AIP-RET interaction in tumorigenesis, we tested the effect of

RET germline missense mutations associated with medullary thyroid carcinoma and MEN2, a dominant

inherited cancer syndrome that affects neuroendocrine organs(18-20). As AIP is also involved in the

tumorigenesis of endocrine tumours, we assessed whether these mutations have a role in impairing AIP-

RET interaction.

Eight missense mutations predisposing to medullary thyroid carcinoma and mapping in the region of

RET that interact with AIP, namely E768D, V778I, L790F, V804M, Y806C, S891A, M918T and

S922F (Fig.6A), were introduced into the RET51-FL construct. The resulting constructs were expressed

in HEK cells with the full-length wild type AIP. Their ability to modify the AIP-RET interaction was

tested by immunoprecipitating the cell lysate with an anti-RET51 antibody and immonoblotting for

AIP. The presence of AIP proteins was detected in the lysate immunoprecipitated with an antibody

13

specific for RET51 recognizing each of the eight mutated forms of RET tested (Fig.6B). These data

clearly showed that none of these eight RET mutations impaired the interaction of RET51 with AIP.

Mutation screening in Growth Hormone (GH)-secreting pituitary adenoma patients–Germline

mutations of AIP were recently identified in somatotropinoma(17, 21, 22). To test the hypothesis that

AIP-RET interaction may have a role in the tumorigenesis of pituitary adenoma and that the loss of

RET protein may predispose to this neoplasia, we decided to examine whether RET mutations occurred

in pituitary adenoma patients. The complete coding sequence of AIP and RET were sequenced in the

pituitary tissues from 28 somatotropinoma patients. Furthermore, the enhancer-like sequence in intron

1of RET containing the non-coding RET variant rs2435357 strongly associated with a loss-of-function

of RET was also evaluated. One already reported mutation in AIP (R304Q) and two unreported variants

in the 3’UTR of RET (c. 1590*G>A; c.1643*G>C) were detected in two different patients. The two

variants influenced neither miRNA binding sites nor mRNA folding, as predicted by in silico analysis.

Furthermore the two alleles of the non-coding regulatory variant of RET were homogeneously

distributed among the patients.

14

DISCUSSION

In this study we identified the co-chaperone AIP as a novel RET-binding partner and showed the

presence of an AIP-RET complex in cell lines of different origin (human embryonic kidney and

neuroblastoma) and in the pituitary gland in vivo. Familial and sporadic pituitary adenoma was recently

associated to germline heterozygous mutations and large genomic deletions in the AIP gene(17, 21-26).

It is tempting to speculate that the interaction between RET and AIP might have a role in pituitary

adenoma tumorigenesis. It has been shown that GDNF, GFRalpha1, and RET mRNA and protein are

expressed both in the human anterior pituitary gland and in somatotropinoma(27, 28) and that the

proto-oncogene RET has a role in regulating apoptosis in somatotrophs both in vitro and in vivo(5).

Indeed, loss of RET function results in somatotroph hyperplasia, which is interpreted to be caused by

the absence of RET-derived pro-death signals(5). The interaction between RET and AIP, shown for the

first time in this work, suggests a possible synergistic activity of RET and AIP in regulating

somatotroph proliferation and tumorigenesis, thus providing a link between pituitary adenoma and RET

dependence receptor activity.

Conversely, we showed that neither RET-associated mutations found in MEN2 nor AIP-associated

mutations found in pituitary adenoma impair the ability of RET to bind AIP. Consequently the

tumorigenic process likely does not result in an impairment of the AIP-RET complex, but rather the

presence of this complex leads to downstream events. So far, little is known about RET signal

transduction leading to apoptosis. AIP may contribute to this process, given its ability to interact with

survivin, a member of the IAP (inhibitor of apoptosis) family (16). We showed that the expression of

RET prevents AIP from binding survivin. Since the binding of AIP to survivin has been demonstrated

to protect survivin from degradation and that the disruption of this complex enhances cell death(16),

RET induced apoptosis might be achieved by preventing the formation of the AIP-survivin complex.

Further work on the functional role of the AIP-RET complex should shed light on the key cellular

processes involved in RET-induced apoptosis. Interestingly, the proapoptotic domain of RET is

responsible for the interaction with AIP. In addition, AIP-RET interaction does not require RET kinase

15

activity or kinase dependent signal transduction such as RET induced cell death(4), further supporting a

role of AIP in modulating RET-induced apoptosis.

Accordingly, RET may be considered a good candidate for pituitary adenoma predisposition. We

therefore screened pituitary adenoma patients for the presence of intragenic RET mutations but did not

find any significant alteration in this gene. This failure does not necessarily exclude RET as a candidate

gene for this form of neoplasia. Actually, despite the sensitivity of the sequencing analysis system we

are aware that it does not detect changes such as large genomic rearrangements or changes in promoter

methylation, which are known mechanisms for tumor suppressor gene inactivation(29, 30). In addition,

our group of patients consists exclusively of sporadic GH-secreting pituitary adenomas with an average

age of 45 years. Selection of patients with a familial history for pituitary adenoma and with an earlier

onset of the disease could potentially increase the chance of finding RET alterations.

Our results provide the basis for unravelling the biological processes underlying AIP-associated

pituitary adenoma predisposition and shedding light on the mechanisms of RET induced cell death.

Further functional studies are needed to completely elucidate the biological mechanisms involving the

RET-AIP complex.

16

Acknowledgments

We thank all the families who have participated in the study and the clinicians who collaborated in

this study. We thank Dr. Giuseppe Gasparre for critical reading and thoughtful discussion of the

manuscript, Dr Lucia Fiammetta Pennisi for her help in some of the experiments and Prof. Kerry

Rhoden for proofreading the manuscript. This work was supported by grant LSHC-CT-2006-037530

“HERMIONE” from the EU to G.R.. M.V. was supported by a University of Turin PhD fellowship.

D.F. was supported by a post-doctoral fellowship from grant HERMIONE.

R.B. was supported by the EC 6th Framework Network of Excellence LIFESPAN (LSHGCT-2007-

036894, and by the Deutsche Forschungsgemeinschaft (CRC746, CRC780), FRISYS (# 03139219,

the Freiburg Initiative in Systems Biology) and FRIAS LIFENET.

17

REFERENCES

1. Trupp M, Raynoschek C, Belluardo N, Ibanez CF 1998 Multiple GPI-anchored receptors

control GDNF-dependent and independent activation of the c-Ret receptor tyrosine kinase.

Mol Cell Neurosci 11:47-63

2. Tahira T, Ishizaka Y, Itoh F, Sugimura T, Nagao M 1990 Characterization of ret proto-

oncogene mRNAs encoding two isoforms of the protein product in a human neuroblastoma

cell line. Oncogene 5:97-102

3. Arighi E, Borrello MG, Sariola H 2005 RET tyrosine kinase signaling in development and

cancer. Cytokine Growth Factor Rev 16:441-467

4. Bordeaux MC, Forcet C, Granger L, Corset V, Bidaud C, Billaud M, Bredesen DE,

Edery P, Mehlen P 2000 The RET proto-oncogene induces apoptosis: a novel mechanism

for Hirschsprung disease. Embo J 19:4056-4063

5. Canibano C, Rodriguez NL, Saez C, Tovar S, Garcia-Lavandeira M, Borrello MG,

Vidal A, Costantini F, Japon M, Dieguez C, Alvarez CV 2007 The dependence receptor

Ret induces apoptosis in somatotrophs through a Pit-1/p53 pathway, preventing tumor

growth. Embo J 26:2015-2028

6. Dirnberger D, Messerschmid M, Baumeister R 2008 An optimized split-ubiquitin cDNA-

library screening system to identify novel interactors of the human Frizzled 1 receptor.

Nucleic Acids Res 36:e37

7. Ma Q, Whitlock JP, Jr. 1996 The aromatic hydrocarbon receptor modulates the Hepa 1c1c7

cell cycle and differentiated state independently of dioxin. Mol Cell Biol 16:2144-2150

8. Carver LA, Bradfield CA 1997 Ligand-dependent interaction of the aryl hydrocarbon

receptor with a novel immunophilin homolog in vivo. J Biol Chem 272:11452-11456

9. Lewis BP, Burge CB, Bartel DP 2005 Conserved seed pairing, often flanked by adenosines,

indicates that thousands of human genes are microRNA targets. Cell 120:15-20

18

10. Ritchie W, Theodule FX, Gautheret D 2008 Mireval: a web tool for simple microRNA

prediction in genome sequences. Bioinformatics 24:1394-1396

11. Enright AJ, John B, Gaul U, Tuschl T, Sander C, Marks DS 2003 MicroRNA targets in

Drosophila. Genome Biol 5:R1

12. Zuker M 2003 Mfold web server for nucleic acid folding and hybridization prediction.

Nucleic Acids Res 31:3406-3415

13. Melillo RM, Cirafici AM, De Falco V, Bellantoni M, Chiappetta G, Fusco A,

Carlomagno F, Picascia A, Tramontano D, Tallini G, Santoro M 2004 The oncogenic

activity of RET point mutants for follicular thyroid cells may account for the occurrence of

papillary thyroid carcinoma in patients affected by familial medullary thyroid carcinoma. Am

J Pathol 165:511-521

14. Tsui-Pierchala BA, Ahrens RC, Crowder RJ, Milbrandt J, Johnson EM, Jr. 2002 The

long and short isoforms of Ret function as independent signaling complexes. J Biol Chem

277:34618-34625

15. Coulpier M AJ, Ibanez CF 2002 Coordinated activation of autophosphorylation sites in the

RET receptor tyrosine kinase: importance of tyrosine 1062 for GDNF mediated neuronal

differentiation and survival. J Biol Chem 277:1991-1999

16. Kang BH, Altieri DC 2006 Regulation of survivin stability by the aryl hydrocarbon receptor-

interacting protein. J Biol Chem 281:24721-24727

17. Vierimaa O, Georgitsi M, Lehtonen R, Vahteristo P, Kokko A, Raitila A, Tuppurainen

K, Ebeling TM, Salmela PI, Paschke R, Gundogdu S, De Menis E, Makinen MJ,

Launonen V, Karhu A, Aaltonen LA 2006 Pituitary adenoma predisposition caused by

germline mutations in the AIP gene. Science 312:1228-1230

18. Eng C 1996 Seminars in medicine of the Beth Israel Hospital, Boston. The RET proto-

oncogene in multiple endocrine neoplasia type 2 and Hirschsprung's disease. N Engl J Med

335:943-951

19

19. Mulligan LM, Kwok JB, Healey CS, Elsdon MJ, Eng C, Gardner E, Love DR, Mole SE,

Moore JK, Papi L. 1993 Germ-line mutations of the RET proto-oncogene in multiple

endocrine neoplasia type 2A. Nature 363:458-460

20. Mulligan LM, Ponder BA 1995 Genetic basis of endocrine disease: multiple endocrine

neoplasia type 2. J Clin Endocrinol Metab 80:1989-1995

21. Georgitsi M, De Menis E, Cannavo S, Makinen MJ, Tuppurainen K, Pauletto P, Curto

L, Weil RJ, Paschke R, Zielinski G, Wasik A, Lubinski J, Vahteristo P, Karhu A,

Aaltonen LA 2008 Aryl hydrocarbon receptor interacting protein (AIP) gene mutation

analysis in children and adolescents with sporadic pituitary adenomas. Clin Endocrinol (Oxf)

69:621-627

22. Georgitsi M, Heliovaara E, Paschke R, Kumar AV, Tischkowitz M, Vierimaa O,

Salmela P, Sane T, De Menis E, Cannavo S, Gundogdu S, Lucassen A, Izatt L, Aylwin S,

Bano G, Hodgson S, Koch CA, Karhu A, Aaltonen LA 2008 Large genomic deletions in

AIP in pituitary adenoma predisposition. J Clin Endocrinol Metab 93:4146-4151

23. Yu R, Bonert V, Saporta I, Raffel LJ, Melmed S 2006 Aryl hydrocarbon receptor

interacting protein variants in sporadic pituitary adenomas. J Clin Endocrinol Metab 91:5126-

5129

24. Toledo RA, Lourenco DM, Jr., Liberman B, Cunha-Neto MB, Cavalcanti MG, Moyses

CB, Toledo SP, Dahia PL 2007 Germline mutation in the aryl hydrocarbon receptor

interacting protein gene in familial somatotropinoma. J Clin Endocrinol Metab 92:1934-1937

25. Leontiou CA, Gueorguiev M, van der Spuy J, Quinton R, Lolli F, Hassan S, Chahal HS,

Igreja SC, Jordan S, Rowe J, Stolbrink M, Christian HC, Wray J, Bishop-Bailey D,

Berney DM, Wass JA, Popovic V, Ribeiro-Oliveira A, Jr., Gadelha MR, Monson JP,

Akker SA, Davis JR, Clayton RN, Yoshimoto K, Iwata T, Matsuno A, Eguchi K, Musat

M, Flanagan D, Peters G, Bolger GB, Chapple JP, Frohman LA, Grossman AB,

Korbonits M 2008 The role of the aryl hydrocarbon receptor-interacting protein gene in

familial and sporadic pituitary adenomas. J Clin Endocrinol Metab 93:2390-2401

20

26. Daly AF, Vanbellinghen JF, Khoo SK, Jaffrain-Rea ML, Naves LA, Guitelman MA,

Murat A, Emy P, Gimenez-Roqueplo AP, Tamburrano G, Raverot G, Barlier A, De

Herder W, Penfornis A, Ciccarelli E, Estour B, Lecomte P, Gatta B, Chabre O, Sabate

MI, Bertagna X, Garcia Basavilbaso N, Stalldecker G, Colao A, Ferolla P, Wemeau JL,

Caron P, Sadoul JL, Oneto A, Archambeaud F, Calender A, Sinilnikova O, Montanana

CF, Cavagnini F, Hana V, Solano A, Delettieres D, Luccio-Camelo DC, Basso A,

Rohmer V, Brue T, Bours V, Teh BT, Beckers A 2007 Aryl hydrocarbon receptor-

interacting protein gene mutations in familial isolated pituitary adenomas: analysis in 73

families. J Clin Endocrinol Metab 92:1891-1896

27. Urbano AG, Suarez-Penaranda JM, Dieguez C, Alvarez CV 2000 GDNF and RET-gene

expression in anterior pituitary-cell types. Endocrinology 141:1893-1896

28. Japon MA, Urbano AG, Saez C, Segura DI, Cerro AL, Dieguez C, Alvarez CV 2002

Glial-derived neurotropic factor and RET gene expression in normal human anterior pituitary

cell types and in pituitary tumors. J Clin Endocrinol Metab 87:1879-1884

29. Xing M 2007 Gene methylation in thyroid tumorigenesis. Endocrinology 148:948-953

30. Segev DL, Umbricht C, Zeiger MA 2003 Molecular pathogenesis of thyroid cancer. Surg

Oncol 12:69-90

21

FIGURE LEGEND

Fig. 1. Schematic structure of RET and RET deleted constructs used for mapping AIP interaction. S,

signal sequence; CAD, cadherin-related motif; CYS, cysteine-rich region; TM, transmembrane

domain; TK, tyrosine kinase domain; 51-SP, specific region of RET long isoform; 9-SP, specific

region of RET short isoform.

Fig. 2. A) Whole cell lysates prepared from HEK293 cells co-transfected with pcDNA3.1-AIP-

myc/His (lane 3) or pcDNA3.1myc/His (lane 1) and RET51-FL. B) Immunoprecipitation was carried

out with anti-RET9 antibody using lysates prepared from HEK 293 cells co-transfected with

pcDNA3.1-AIP-myc/His (lane 1) or pcDNA3.1myc/His (lane 3) and RET9-FL. C) Whole cell lysates

prepared from Neuro2A and SH-SY5Y and immunoblotted for anti-RET antibody and anti-AIP

antibody. D) Whole cell lysates prepared from Neuro2A cells co-transfected with pcDNA3.1-AIP-

myc/His (lane 1) or pcDNA3.1myc/His (lane 3) and RET51-FL and from Neuro2A not transfected

(lane 5). D) HEK293 cells co-transfected with pcDNA3.1-AIP-myc/His and RET51-FL (lane 1) or

RET51-D707N (lane 3) . E) HEK293 cells co-transfected with pcDNA3.1-AIP (lane 3 and 5) or

pcDNA3.1(+) (lane 1) and RET-PRO (lane 1 and 5) or pcDNA3.1myc/His (lane 3). F) HEK293 cells

co-transfected with pcDNA3.1-AIP-myc/His (lane 1 and 5) or pcDNA3.1myc/His (lane 3 and 7) and

RET-IUX (lane 5 and 7) or RET-TK (lane 1 and 3). G) HEK293 cells co-transfected with pcDNA3.1-

AIP-myc/His (lane 1 and 3) or pcDNA3.1myc/His (lane 5 and 7) and RET-IUX (lane 1 and 5) or

RET-TK (lane 3 and 7). All the western blot were performed with anti-myc antibody except for (F)

and (C) where immunoblotting was carried out with anti- RET antibody.

Fig. 3. A) Protein extracted from rat pituitary gland were immunoblotted for either anti-AIP or anti-

RET antibody. B) Whole lysate prepared from rat pituitary gland were immunoprecipitated with

different antibodies. A fifth of the lysate immunoprecipitated with anti-AIP antibody was loaded in

Lane1. Whole lysate immunoprecipitated with anti-RET antibody and for a not related

22

immunoglobulin were loaded in Lane2 and Lane 3, respectively. Immunoprecipitates were

immunoblotted with anti-AIP antibody.

Fig. 4. A) Immunoprecipitation was carried out with anti-RET51 antibody using lysates prepared

from HEK 293 cells co-transfected with pcDNA3.1-AIP-myc/His (lane 1 and 3) or

pcDNA3.1myc/His (lane 5) and RET51-FL (lane 1 and 5) or RET51-DK (lane 3). B)

Immunoprecipitation was carried out with anti-survivin antibody using lysates prepared from HEK

293 cells transfected with pcDNA3.1-AIP-myc/His alone (lane 1) or together with RET51-FL (lane

3). C) Hypothetical mechanism of RET induced apoptosis. When RET is not expressed, AIP directly

bind survivin, preventing its degradation thus protecting cell from apoptosis. The presence of RET

abolished the ability of AIP to interact with survivin, likely leading to survivin degradation and

consequent cell death.

Fig. 5. A) Schematic figure of AIP. FK506-binding protein (FKBP)–homology region and

tetratricopeptide repeats (TPRs) are shown. The regions necessary for AHR and HSP90 interaction

are shown with black lines and the exon boundaries of the AIP gene are marked with dashed lines.

AIP mutations tested for their ability to impair the interaction with RET are indicated by black

triangles. B) HEK293 cells co-transfected with pcDNA3.1-AIP-myc/His (lane 1 and 5) or

pcDNA3.1myc/His (lane 3) or the mutated constructs of AIP (lane 7-18) and RET51-FL (lane 1 and

3) or the empty vector (Ω) (lane 5). C) HEK293 cells co-transfected with pcDNA3.1-AIP (lane 1 and

5) or pcDNA3.1(+) (lane 3) or the mutated constructs of AIP (lane 7-18) and RET-PRO (lane 1 and 3)

or pcDNA3.1myc/His (lane 5). Mutations tested: lane 7: R16H; lane 9: K241E; lane 11: R271W; lane

13: E293G; lane 15: A299V; lane 18: R304Q. All immunoblottings were performed with anti-myc

antibody.

Fig. 6. A) Schematic structure of RET. The position of the RET missense mutations mapping to the

AIP interacting region are indicated by arrows. B) HEK293 cells co-transfected with pcDNA3.1-AIP-

23

myc/His (lane 1-18) and RET51-FL (lane 1) or the mutated constructs of RET (lane 3-18). Mutations

tested: lane 3: E768D; lane 5: V778I; lane 7: L790F; lane 9: V804M; lane 11: Y806C; lane 13:

S891A; lane 15: M918T; lane 18: S922F. All immunoblottings were performed with anti-myc

antibody.

707

TK TK

1017

myc

1

S CAD CYS TKTM TK

999V5

1

S CAD CYS TM

725

V5

S CAD CYS TKTM TK 51-SP

1 1114

S CAD CYS TKTM TK 9-SP

1 1072RET-9 FL

RET-51 FL

RET-TK

RET-IUX

RET-PRO

Figure 1

Figure 2

A B

D

E

F

IP α

-RET51

WB α -myc

myc emptyRET9

AIPRET9

AIPRET51

myc emptyRET51

myc emptyRET51

AIPRET51

AIPIUX

myc emptyTK

myc emptyIUX

AIPTK

AIPIUX

myc emptyTK

myc emptyIUX

AIPTK

C

WB α-AIP

WB α-RET

N2A SY

GAIP

myc

AIP

PRO

3.1PRO

WB α -myc

mock

IP α

-RET51

input

input

IP α

-RET9

IP α

-RET9

input

input

IP α

-RET51

IP α

-RET51

input

input

IP α

-RET51

input

IP α

-myc

WB α-RET

IP α

-myc

input

input

IP α

-myc

input

IP α

-myc

input

IP α

-RET

WB α-myc

IP α

-RET

input

input

IP α

-RET

input

IP α

-RET

input

IP α

-AIP

WB α-myc

IP α

-AIP

input

input

IP α

-AIP

input

WB α -myc

Figure 3

WB anti AIP

IP a

nti-AIP

IP a

nti-RET

IP u

nrela

ted Ig

B

A

36 KDa

36 KDa

120 KDa

WB anti AIP WB antiRET

Figure 4

survivin

myc emptyRET51

AIPRET51

AIPK758R

AIPRET51 AIP

B

A

36 KDa

36 KDa

C

survivin

AIP survivin

AIP

RE

TIP

α-s

urviv

in

WB α-AIP

IP α

-surv

ivin

input

input

IP α

-RET51WB α-myc

IP α

-RET51

input

input

IP α

-RET51

input

Figure 5

R16H

AHR

HSP90

K241E

E293G

R271W

A2

99V

TP

RFKBP

TP

R

TP

R

R304Q

A

B

C

WB anti-myc

WB anti-myc

A299VRET51

E293GRET51

R304QRET51

R16H

RET51

R271W

RET51

K241E

RET51

AIPΩ

myc empty

RET51

AIPRET51

AIPmyc

3.1PRO

AIPPRO

A299V

PRO

E293G

PRO

R304Q

PRO

R16HPRO

R271WPRO

K241EPRO

36 KDa

36 KDa

36 KDa

36 KDa

36 KDa

36 KDa

IP α

-RET51

IP α

-RET51

input

input

IP α

-RET51

input

IP α

-RET51

IP α

-RET51

input

input

IP α

-RET51

input

IP α

-RET51

IP α

-RET51

input

input

IP α

-RET51

input

IP α

-AIP

IP α

-AIP

input

input

IP α

-AIP

input

IP α

-AIP

IP α

-AIP

input

input

IP α

-AIP

input

IP α

-AIP

IP α

-AIP

input

input

IP α

-AIP

input

Figure 6

A

B

S CAD CYS TKTM TK 51-SP

1 1114

L790F Y806C

V804MV778I

E768D M918T

S891A S922F

WB anti-myc

AIPL790F

AIP Y806C

AIP V804M

AIPV778I

AIPE768D

AIP

RET51

AIP

M918T

AIP

S891AAIP

S922F

36 KDa

36 KDa

36 KDa

IP α

-RET51

IP α

-RET51

input

input

IP α

-RET51

input

IP α

-RET51

IP α

-RET51

input

input

IP α

-RET51

input

IP α

-RET51

IP α

-RET51

input

input

IP α

-RET51

input