Anne 2006-2007 UNIVERSITE TOULOUSE III PAUL SABATIER Ecole doctorale Biologie-Sant-Biotechnologie

THESE
DOCTEUR DE LUNIVERSITE TOULOUSE III
Discipline : Cancrologie
Prsente et soutenue par

Pour obtenir le grade de

Nas PRADE-HOUDELLIER
Le 30 Novembre 2007

Rgulation de la tlomrase dans les cellules hmatopotiques normales et leucmiques

Directeur de thse : Dr. Vronique DE MAS

JURY M. le Professeur Guy LAURENT M. le Professeur Ali BETTAIEB M. le Docteur Hamid MORJANI M. le Professeur Catherine MULLER Prsident du jury Rapporteur Rapporteur Examinateur

TABLE DES MATIERES

ABBREVIATIONS .................................................................................................4 INTRODUCTION ................................................................................................... 5

PARTIE 1 : ETUDE BIBLIOGRAPHIQUE ....................... 7

I. Les tlomres..................................................................................................... 8
1. Dfinition et structure ................................................................................ 8 2. Techniques de mesure de la longueur des tlomres ...................... 9 2.1. Technique TRF ............................................................................................ 9 2.2. Q-FISH ........................................................................................................ 10 2.3. Flow-FISH ................................................................................................... 10

II. La Tlomrase : gnralits et structure .............................................. 11

1. La sous-unit ribonuclique ...................................................................... 11 2. La sous-unit catalytique ........................................................................... 13 3. Techniques de mesure de lactivit de la tlomrase ....................... 15 3.1. TRAP semi-quantitatif................................................................................. 15 3.2. TRAP quantitatif.......................................................................................... 16

III. Mcanismes de rgulation de la tlomrase .................................... 17

1. Les mcanismes de rgulation de hTERT ............................................ 17
1.1. 1.2. 1.3. 1.4. Rgulation transcriptionnelle...................................................................... 17 Rgulation pigntique de la transcription ............................................... 19 Rgulation par pissage alternatif............................................................... 20 Rgulation post-traductionnelle.................................................................. 20 1.4.1. Translocation au noyau..................................................................... 21 1.4.2. Repliement et assemblage ................................................................ 22 1.4.3. Phosphorylation ................................................................................ 22 2. Rgulation de la sous-unit ribonuclique ........................................... 23 2.1. Rgulation transcriptionnelle et post-transcriptionnelle ............................. 23 2.2. Stabilit de hTR........................................................................................... 24

3. Les signaux intra et extracellulaires conduisant la rgulation de la tlomrase ....................................................................... 24

3.1. Les hormones .............................................................................................. 24 3.2. Les facteurs de croissance........................................................................... 25 3.2. Les cytokines............................................................................................... 25

IV. Rles de la tlomrase .............................................................................. 27

1. Maintien de la longueur des tlomres .................................................. 27 2. 3. 4. 5. 6. 7.
1.1. La snescence rplicative............................................................................ 27 1.2. La snescence prmature induite par le stress .......................................... 28 Prolifration .................................................................................................... 29 Protection contre lapoptose ..................................................................... 29 Rparation de lADN ................................................................................... 30 Diffrenciation ............................................................................................... 31 Modles murins de modulation de la tlomrase .............................. 32 Tlomrase et cancer ................................................................................... 33

V. Tlomrase et hmatopose.................................................................... 36
1. Lhmatopose .............................................................................................. 36 2. La tlomrase et les tlomres dans les cellules hmatopotiques .......................................................................................... 39

2.1. Tlomrase et diffrenciation hmatopotique.......................................... 39 2.2. Statut des tlomres dans les cellules hmatopotiques ............................ 40 2.3. Pathologies constitutionnelles ou acquises lies la tlomrase touchant lhmatopose.......................................................... 41 3. Tlomrase et hmopathies malignes..................................................... 41 3.1. Leucmies aigus........................................................................................ 42 3.2. Syndromes mylodysplasiques................................................................... 43 4. Rle de la tlomrase dans lhmatopose ? ...................................... 44

PARTIE 2 : RESULTATS .................................................................... 47

I. Objectifs du travail ......................................................................................... 48 II. Rsultats 1 : Rgulation de lexpression de la tlomrase par le TNF- dans les cellules mylodes humaines normales et leucmiques ............................................................................. 50
2

III. Rsultats 2 : Rle et rgulation de la tlomrase dans les progniteurs rythrodes humains ...................................... 59

DISCUSSION ET PERSPECTIVES ................................................................ 73 REFERENCES ......................................................................................................... 79 ANNEXES .................................................................................................................. 99

ABBREVIATIONS
ADN ALT ARN BFU-E BrdU CFU-E CFU-G CFU-GEMM CFU-GM CFU-M CSH DN-hTERT ELISA EPO EPOR Flow-FISH FLT3-L FRET GM-CSF GPA hTER ou hTR ou hTERC hTERT IL IRES JAK2 LAM MAPK mTERC mTERT PI3K Q-FISH RTL RT-PCR SCF siRNA SMD snoRNA STAT TGFTNFTNFR TRAP WT Acide desoxyribonuclique Alternative Lenghtening of Telomeres Acide ribonuclique Burst Forming Unit Erythroid Bromo doxyUridine Colony Forming Unit Erythroid Colony Forming Unit Granulocytic Colony Forming Unit Colony Forming Unit Granulo/Monocytic Colony Forming Unit Monocytic Cellule Souche Hematopotique Dominant Ngatif de la tlomrase Enzymatic Linked Immuno Sorbent Assay Erythropotine Erythropoietin Receptor Flow Cytometry Fluorescence In Situ Hybridation FMS-like tyrosine kinase 3 ligand Fluorescence Resonance Energy Transfer Granulo/Monocyte Colony Stimulating Factor Glycophorin A human Telomerase RNA human Telomerase Reverse Transcriptase Interleukine Internal Ribosome Entry Site Janus kinase 2 Leucmie aigu mylode Mitogen Activated Protein Kinase mouse Telomerase RNA mouse Telomerase Reverse Transcriptase Phosphatidyl Inositol 3 Kinase Quantitative Fluorescence In Situ Hybridation Relative Telomere Length Reverse Transcriptase Polymerase Chain Reaction Stem Cell Factor small interfering RNA Syndrome Mylodysplasique small nucleolar ribonucleic acid Signal Transducer and Activator of Transcription Transforming Growth Factor Tumor Necrosis Factor TNF Receptor Telomere Repeat Amplification Protocol Wild Type (forme sauvage)

INTRODUCTION

Chez les organismes eucaryotes, on trouve lextrmit des chromosomes une squence nuclotidique rpte non codante associe de nombreuses protines, appele tlomre. Ces squences tlomriques permettent non seulement de protger le gnome dune perte dinformation due un raccourcissement progressif des chromosomes chaque division cellulaire, mais aussi de protger les extrmits naturelles des chromosomes contre les fusions bout bout et les recombinaisons dues aux systmes de rparations des dommages de lADN. Les tlomres, en se raccourcissant chaque division, sont aussi impliqus dans les processus de vieillissement cellulaire. En effet, lorsque les tlomres atteignent une taille critique, la cellule entre en snescence et arrte de prolifrer. La cellule peut mettre en uvre deux stratgies diffrentes afin dviter lattrition tlomrique, le mcanisme dALT (Alternative Lenghtening of Telomere) par recombinaison, ou lexpression de la tlomrase. La tlomrase est une transcriptase inverse constitue dune sous unit catalytique hTERT et dune sous unit ribonuclique hTR. Sa fonction primordiale est de compenser lrosion des tlomres lors de la prolifration cellulaire en rajoutant des squences nuclotidiques spcifiques lextrmit 5 des rgions tlomriques. Rcemment, dautres fonctions ont t attribues la tlomrase, notamment dans la prolifration, la protection contre lapoptose, la diffrenciation, ou la rparation de lADN, ce qui fait de cette enzyme un acteur majeur dans de nombreux domaines. Le niveau dexpression dhTERT dans la cellule (facteur limitant de lactivit tlomrase), est soumis un grand nombre de rgulation tant transcriptionnel que post-transcriptionnel. On retrouve une activit tlomrase leve dans plus de 80% des cancers, et mme sil est vrai que les cellules physiologiques ont gnralement une activit tlomrase pratiquement inexistante, il existe certaines exceptions, notamment dans les cellules assurant un renouvellement rapide de certains tissus : cest le cas de lpithlium digestif et surtout de lhmatopose. Lhmatopose est le processus permettant lobtention de toutes les cellules sanguines partir de cellules souches multipotentes prsentes dans la moelle osseuse. Ce processus est rgul par de nombreux facteurs de croissance ou cytokines produites par les

cellules de lenvironnement mdullaire, pouvant avoir un effet soit activateur, soit inhibiteur. Lactivit de la tlomrase dans les cellules hmatopotiques varie de faon importante tout au long du processus de diffrenciation. Ainsi le niveau dactivit tlomrase dans les cellules souches hmatopotique est trs faible, puis celui-ci augmente dans les progniteurs des diffrentes lignes sanguines pour revenir un taux basal dans les cellules matures circulantes. Par ailleurs, hTERT est retrouve surexprime dans prs de 80% des cas de leucmies aigus. Lensemble de ces observations suggre que, non seulement la tlomrase semble jouer un rle prpondrant dans le processus hmatopotique, mais que dautre part, son expression pourrait tre rgule par les facteurs de croissance et cytokines hmatopotiques permettant lobtention des cellules sanguines. On peut de mme supposer que toute altration de ces voies de rgulation du niveau dactivit tlomrase, pourrait entraner une hmatopose inefficace.

PARTIE 1 : ETUDE BIBLIOGRAPHIQUE

ETUDE BIBLIOGRAPHIQUE

I. Les tlomres
1. Dfinition et structure
Chez les organismes eucaryotes qui ont la particularit davoir des chromosomes linaires, on observe chaque rplication de lADN, une perte de plusieurs dizaines de nuclotides au niveau de lextrmit 3 du chromosome. Afin dviter cette perte dinformation gntique, on trouve lextrmit des chromosomes une squence nuclotidique rpte non codante associe de nombreuses protines, appele tlomre. Ces squences tlomriques permettent non seulement de protger le gnome dune perte dinformation due un raccourcissement progressif des chromosomes, mais aussi de protger les extrmits naturelles des chromosomes contre les fusions bout bout et les recombinaisons dues aux systmes de rparations des dommages de lADN. Le concept de tlomre a t voqu pour la premire fois par les travaux de Mller et McClintock au dbut des annes 40 (Mller 1940; McClintock 1941), puis en ce qui concerne leur implication dans la rplication de lADN, par Olovnikov et Watson au cours des annes 70 (Watson 1972; Olovnikov 1973). Les tlomres sont constitus dune rptition de courtes squences nuclotidiques dont la longueur totale peut aller de 20 plusieurs milliers de paires de bases. LADN tlomrique des vertbrs est ainsi compos dune longue squence double brin (contenant les rptitions 5-TTAGGG-3 chez lhomme ou la souris par exemple), termin par une extrmit sortante simple brin en 3 contenant un nombre variable de squences tlomriques (Blackburn, Greider et al. 1989). Chez lhomme et la souris entre autres, il a t montr que lextrmit simple brin pouvait shybrider au sein de la squence double brin afin de former une grande boucle appele T-loop (Griffith, Comeau et al. 1999). On sait de plus que de nombreuses protines sont associes aux tlomres dans cette boucle permettant la formation dun complexe tlomrique servant de coiffe pour le tlomre en protgeant son extrmit (Figure 1). Parmi les protines se liant directement lADN tlomrique, on retrouve les protines TRF1 et TRF2 (Telomere Repeat binding Factor 1 et 8

2) (Broccoli, Smogorzewska et al. 1997), POT1 (Protection of Telomere 1) et hnRNPA1 (heterogenous nuclear ribonucleoprotein A1) (Baumann, Podell et al. 2002). Dautres protines viennent ensuite se lier aux premires dont TIN2 (TRF1-interacting nuclear factor 2) (Kim, Kaminker et al. 1999), RAP1 (Repressor/activator protein 1) (Li, Oestreich et al. 2000), TANK1 et 2 (tankyrase) (Kaminker, Kim et al. 2001), PinX1 (Zhou and Lu 2001) et certaines protines de la rparation (complexe DNA-PK et complexe Rad50/Mre11/NBS1 (RMN)) (Song, Jung et al. 2000; Wu, Lee et al. 2000).

RMN

Ku RAP1

PinX 1

TRF1 TRF2 POT1

T-loop
hnRNPA1

TIN2

Nuclosomes
Figure 1 : Complexe tlomrique (daprs (Blackburn 2001))

TANK

2. Techniques de mesure de la longueur des tlomres

2.1. Technique TRF (Telomere Restriction Fragments) Cette technique permet la mesure de la taille moyenne des tlomres dune population donne par Southern blot (Vaziri, Schachter et al. 1993). Aprs digestion de lADN gnomique par les enzymes de restriction Hinf I et Rsa I qui permettent de prserver les rptitions tlomriques TTAGGG, les produits de digestion sont spars sur gel dagarose, puis rvls par autoradiographie ou chemiluminescence. Les fragments tlomriques apparaissent sous forme de smear et leur longueur moyenne est calcule en comparant au marqueur de taille de lADN dpos sur le gel. Cette technique permet

donc dobtenir la longueur moyenne des tlomres dune population cellulaire, mais la quantification reste imprcise. 2.2. Q-FISH (Quantitative-Fluorescence in situ hybridation) Cette technique utilise des sondes dacides nucliques peptidiques (PNA Peptide Nucleic Acid) de squences (CCCTAA)3 marques par un fluorochrome. Les sondes vont reconnatre les extrmits tlomriques et shybrider. La lecture grce un microscope fluorescence des cellules en mtaphase va permettre de mesurer lintensit de fluorescence au niveau de chaque tlomre, puis en comparant avec la fluorescence de plasmides calibrateurs, de calculer prcisment la longueur des tlomres (Poon, Martens et al. 1999). Cette technique prsente donc lavantage de permettre de mesurer la longueur des tlomres au niveau de chaque chromosome et non une moyenne de la population. 2.3. Flow-FISH Le Flow-FISH est une adaptation de la technique de Q-FISH la cytomtrie en flux (Rufer, Dragowska et al. 1998). Les sondes de PNA sont hybrides sur les cellules en suspension, et la lecture au cytomtre en flux permet dobtenir une mesure de la moyenne de fluorescence dans lchantillon de cellules. Un double marquage liodure de propidium doit tre ralis afin de slectionner uniquement les cellules en phase G0/G1 du cycle cellulaire (une seul copie dADN donc quantit de tlomres constante). La quantification peut tre ralise en comparant la fluorescence de lchantillon avec celle dune ligne cellulaire dont la longueur des tlomres reste constante que lon utilise comme talon interne. Les rsultats sont alors obtenus en longueur relative ou RTL (Relative Telomere Lenght).

10

II. La Tlomrase : gnralits et structure

Comme nous lavons vu, chaque rplication de lADN implique une perte dau moins une dizaine de nuclotides au niveau des tlomres, lextrmit des chromosomes. Les cellules immortelles ou cancreuses doivent donc possder un mcanisme permettant dviter lattrition totale des tlomres. Il faudra attendre 1985, pour que la notion de tlomrase soit voque. En effet, Greider et Blackburn montrent que des extraits cellulaires de Tetrahymena (famille des cilis) contiennent une activit permettant lajout de squences rptes TTGGGG la suite damorces tlomriques (Greider and Blackburn 1985). Lenzyme implique dans ce processus est la tlomrase. La tlomrase est une enzyme doue dune activit transcriptase inverse, qui permet de rallonger les extrmits tlomriques par addition de squences spcifiques rptes, et ainsi de limiter leur rosion au cours des divisions cellulaires. La tlomrase humaine est une ribonucloprotine compose dune sous-unit catalytique activit transcriptase inverse, hTERT (human TElomerase Reverse Transcriptase), et dune sous-unit ribonuclique compose dune matrice dARN, hTR (ou hTER pour human Telomerase RNA) ou hTERC (human telomerase RNA component), permettant laddition de squences TTAGGG lextrmit 3 du tlomre.

1. La sous-unit ribonuclique
A la suite de leurs travaux de dcouverte de la tlomrase chez le cili Tetrahymena, lquipe de Blackburn montrent que cette enzyme est dpendante dun ARN pour son activit (Greider and Blackburn 1989; Yu and Blackburn 1991). Cet ARN, repli dans une structure tridimensionnelle complexe, sert de matrice pour laddition des squences tlomriques par la sous-unit catalytique de la tlomrase. La squence primaire de cet ARN est trs peu conserve entre les diffrentes espces. Mme au sein des vertbrs on trouve une grande divergence de squences. Par exemple, il ny a que 65% didentit entre les sous-units dARN de la tlomrase humaine et murine (mTERC). En revanche, la structure secondaire est paradoxalement bien conserve (Chen, Blasco et al. 2000).

11

Domaine CR4-CR5

Domaine pseudoknot

TBE Boite H/ACA

Matrice

Domaine CR7

Figure 2 : A) Structure de la sous-unit ribonuclique de la tlomrase humaine. B) Proposition de structure tridimensionnelle du domaine pseudoknot (Theimer, Blois et al. 2005).

Chez lhomme, la sous-unit ribonuclique, appele hTR (human Telomerase RNA), est compose dun ARN intgral de 451 nuclotides, cod par un gne situ en 3q26. Une tude phylogntique des ARN de tlomrase chez de nombreuses espces de vertbrs a permis de proposer un modle de structure secondaire pour lARN de la tlomrase humaine (Figure 2) (Chen, Blasco et al. 2000). Cette tude a permis de mettre en vidence lexistence de 8 rgions conserves, CR1 CR8, dans la squence des ARN de tlomrase, qui sorganisent en 4 structures conserves spares par des rgions apparies hypervariables. La rgion CR1, proche de lextrmit 5, contient la squence matrice, 5CUAAUCCU-3 chez lhomme. Le domaine pseudoknot contient les rgions CR2 et CR3, immdiatement suivies par une squence appele TBE (Template boundary element). Il sorganise en une structure complexe compacte dont le rle, bien quindispensable lactivit tlomrase, nest pas encore bien dfini. Le domaine CR4-CR5, form dune structure en tige et boucle, permet une interaction avec la sous-unit catalytique hTERT. Le domaine boite H/ACA inclut les rgions CR6 et CR8, et prsente une structure similaire

12

 celle retrouve dans les snoRNAs (Mitchell, Cheng et al. 1999). Enfin, le domaine CR7 nest spcifique quaux ARN de tlomrase des vertbrs. Lquipe de Blackburn a dmontr quen dehors de la squence matrice, seuls les domaines pseudoknot et CR4CR5 sont indispensables la reconstitution dune tlomrase catalytiquement active in vitro (Ly, Blackburn et al. 2003).

2. La sous-unit catalytique
Le gne codant pour la sous-unit catalytique de la tlomrase, TERT, a t tout dabord clon chez la levure (Saccharomyces cerevisiae), et un protozoaire cili (Euplotes aediculatus) (Lendvay, Morris et al. 1996; Lingner and Cech 1996). La comparaison des squences de ces protines avec les banques de donnes a permis de mettre en vidence la prsence de 7 domaines conservs caractristiques des transcriptases inverses (Lingner, Hughes et al. 1997)) (Figure 3). De plus, il apparat que les TERTs font parties dune grande famille de protines caractrises par une structure en doigts-paume-pouce (Nugent and Lundblad 1998). Cependant il existe plusieurs diffrences de structure entre les protines de la famille TERT et les transcriptases inverses classiques. On note en effet la prsence dun important domaine N-terminal (NTE), et dun motif T ncessaire lactivit enzymatique. De plus, les motifs A et B situs au niveau de la structure doigtspaume-pouce , sont spars par un plus grand nombre de nuclotides, ce qui suggre une fonction diffrente de cette rgion par rapport aux autres transcriptases inverses (Nugent and Lundblad 1998). Ces diffrences de structure refltent une diffrence dactivit catalytique, en effet, contrairement aux transcriptases inverses conventionnelles, les TERTs ne sont pas capables de transcrire de longues squences dARN, elles ne copient que quelques nuclotides puis rptent lopration aprs translocation de la matrice dARN (processivit) afin dobtenir des squences tlomriques rptes. Plusieurs conditions sont ncessaires cette activit enzymatique : lintgrit du site catalytique, la liaison la matrice dARN, et la dimrisation ou loligomrisation de la protine.

13

doigts-paume

pouce

Figure 3 : Structure de la sous-unit catalytique de la tlomrase humaine hTERT, comparaison avec la transcriptase inverse du virus HIV. (Autexier and Lue 2006)

Chez lhomme, le gne de la sous-unit catalytique hTERT est situ en 5p15.33, et code pour une protine de 1132 acides amins de 127kDa (Cong, Wen et al. 1999). Comme les autres TERTs, elle est compose dun long domaine N-terminal (NTE) de 350 acides amins, dun motif T, des 7 domaines caractristiques aux transcriptases inverses et dun domaine C-terminal (CTE) de 218 acides amins (Figure 3). Chacun de ces domaines est indispensable lactivit catalytique de lenzyme : le domaine N-terminal permet linteraction avec la sous-unit ribonuclique travers les motifs GQ, CP et QFP, ainsi que loligomrisation grce la partie Cterminale de ce domaine, les motifs T et 1 E situs dans la partie centrale de la protine comprennent le site catalytique de lenzyme, le domaine C-terminal joue un rle la fois dans la processivit (partie la plus proche du site catalytique), loligomrisation, et la localisation cellulaire (Revue par (Autexier and Lue 2006)).

14

3. Techniques de mesure de lactivit tlomrase

La dtection de lactivit de la tlomrase dans des chantillons biologiques peut tre ralise grce la mthode du TRAP (Telomere Repeat Amplification Protocol). Cette technique est compose de deux phases, la premire permet llongation in vitro dun fragment tlomrique synthtique par la tlomrase contenue dans lextrait cellulaire, la deuxime permet la dtection de ces produits dlongation par amplification par PCR (Kim, Piatyszek et al. 1994). Le principe initial de la technique TRAP consiste en lutilisation dune amorce synthtique (TS) mimant une extrmit tlomrique et dune seconde amorce (CX) permettant lamplification par PCR des produits dlongation. La tlomrase contenue dans lextrait cellulaire va ainsi, en prsence de dNTP, ajouter des rptitions tlomriques lextrmit 3 de lamorce TS : 5-AATCCGTCGACGAGAGTT-3 pendant une phase dincubation de 20 30 min 25C. Par la suite la prsence de Taq polymrase et de lamorce CX : 5-AATCCCATTCCCATTCCCATTCCC-3 permet lamplification par PCR des produits forms par la tlomrase. La migration de ces produits damplification sur gel de polyacrylamide en prsence dthydium bromide, permet de distinguer des bandes successives spares de 6 nuclotides (aspect en barreaux dchelle) correspondant aux produits de PCR forms aprs lajout dun nombre distinct de rptitions tlomriques lamorce TS. Il existe plusieurs techniques de dtection de ces produits damplification qui vont permettre une mesure semi-quantitative ou quantitative de lactivit tlomrase initiale. 3.1. TRAP semi-quantitatif Lajout dun standard interne dans le milieu ractionnel permet dobtenir une mesure semi-quantitative de lactivit tlomrase. Ce standard est en effet form dune squence nuclotidique flanque en 5 et 3 des squences complmentaires aux amorces TS et CX et sera donc amplifi de la mme manire que le produit de la raction de TRAP lors de la PCR. Ce standard tant ajout en quantit quivalente dans le milieu ractionnel de chaque chantillon, il permettra non seulement de contrler labsence dinhibiteurs de Taq polymrase dans lextrait cellulaire, mais aussi de normaliser le signal issu de lamplification des produits dlongation de la tlomrase.

15

Plusieurs techniques permettent par la suite de dtecter les produits damplification de la PCR. Un test utilisant la mthode de transfert dnergie en fluorescence (FRET) permet, grce lutilisation dune amorce-sonde dhybridation en pingle cheveux marque par un fluorochrome, une semi-quantification des produits de PCR (TRAPEZEXL, Chemicon International) (Uehara, Nardone et al. 1999). Les produits de PCR peuvent aussi tre dtects par ELISA, en marquant lamorce TS avec la biotine. Aprs immobilisation en plaque de microtitration sur laquelle de la streptavidine (qui fixe la biotine) a t fixe, les produits de PCR sont dtects grce une sonde doligonuclotides couple la proxidase (TeloTAGGG Telomerase PCR-ELISA kit, Roche) (Wu, Hruszkewycz et al. 2000). 3.2. TRAP quantitatif Plus rcemment, une technique de TRAP quantitatif a pu tre dvelopp grce lutilisation de la PCR quantitative en temps rel (Hou, Xu et al. 2001). Ce test utilise le mme principe que le TRAP classique, mais la phase damplification sera ralise en PCR temps rel qui permettra la dtection chaque cycle des produits damplification grce la prsence de SybrGreen I qui reconnat les fragments dADN double brin. La ralisation dune gamme talon obtenue partir de dilutions successives dun chantillon de tlomrase quantifi, permettra la quantification prcise de lactivit tlomrase contenue dans lchantillon de dpart.

16

III. Mcanismes de rgulation de la tlomrase

La plupart des cellules somatiques humaines prsentent une activit tlomrase pratiquement nulle. Seules les cellules haute capacit prolifrative, telles que certaines cellules souches de la peau ou de lintestin, quelques populations lymphocytaires et les cellules souches hmatopotiques, ont une activit tlomrase dtectable. Chez lhomme, la sous-unit ribonuclique de la tlomrase est exprime de faon ubiquitaire dans la quasi-totalit des cellules, et il existe donc peu de mcanismes conduisant sa rgulation. En revanche, la sous-unit catalytique est fortement rgule dune cellule lautre. Plusieurs quipes ont dailleurs montr que lexpression de hTERT tait le facteur limitant de lactivit tlomrase et que son expression ectopique permettait de rtablir une activit tlomrase dans des cellules qui en taient dpourvues (Kirkpatrick, Clark et al. 2003).

1. Les mcanismes de rgulation de hTERT

1.1 Rgulation transcriptionnelle

ERE

+ + +/- ERE/Sp1 NF-kB

AP1

MZF-2

WT1

Max/Myc (Mad)

+/-

p53

E2F

Sp1

?
E-box GC-box

Ets

E2F Sp1

- +

Max/Myc (Mad)

+/-

HTERT
E-box

ATG
+1

-2000

-1000

-200

Figure 4 : Rgulation transcriptionnelle du gne hTERT : fixation des diffrents facteurs de transcription sur le promoteur du gne hTERT. ERE : Estrogen responsive element, Sp1 : Stimulating protein 1, MZF-2 : Myeloid zinc finger 2, WT1 : Wilms tumor 1, Ets : E twenty six specific. + : activateur de la transcription, - : rpresseur de la transcription.

17

Le promoteur du gne de la sousunit catalytique de la tlomrase est une squence denviron 1100 paires de bases en amont du site dinitiation de la transcription (Wick, Zubov et al. 1999). De nombreuses tudes ont permis didentifier les facteurs de transcription et les rpresseurs responsables de la rgulation du gne de hTERT, parmi lesquels on retrouve principalement c-Myc/Mad, Sp1, le rcepteur aux estrognes, Ets, NFB, E2F-1, MZF-2 et WT1 (Figure 4). Le promoteur du gne de hTERT contient deux sites de fixation de c-Myc, des Ebox (de squence conserve (CACGTG)), lun juste en amont du site dinitiation de la transcription, lautre environ -200 paires de bases (Greenberg, O' Hagan et al. 1999). En se fixant sur cette squence consensus en association avec la protine Max, c-Myc a un effet dactivateur de la transcription du gne (Wang, Xie et al. 1998; Wu, Grandori et al. 1999). En effet, plusieurs tudes ont montr que la transcription du gne hTERT pouvait tre influence par une rgulation de c-Myc. Par exemple, le cramide induit une inhibition transcriptionnelle de hTERT par protolyse de c-Myc aprs ubiquitination (Ogretmen, Kraveka et al. 2001). En revanche, si lhtrodimre Mad/Max se fixe la place de Myc/Max, la transcription du gne sera inhibe (Zou, Zhang et al. 2005). Ainsi, dans un contexte cellulaire o la tlomrase est trs faiblement exprime comme les cellules somatiques normales, on retrouve une faible expression de c-Myc et une forte expression de Mad1 (Gunes, Lichtsteiner et al. 2000; Oh, Song et al. 2000). La balance entre les quantits de c-Myc et de Mad1 dans la cellule aura donc une influence sur lactivit transcriptionnelle du gne hTERT. De mme que pour c-Myc, le promoteur de hTERT contient plusieurs sites potentiels de fixation de Sp1 (Stimulating Protein 1), les GC-box (Wick, Zubov et al. 1999). Ce facteur de transcription agit en coopration avec c-Myc pour trans-activer le gne hTERT, en effet lorsque le promoteur est mut sur les sites de fixation de Sp1, la transcription de hTERT induite par c-Myc est beaucoup plus faible (Kyo, Takakura et al. 2000). Plus rcemment, il a t montr que dans des cellules dadnocarcinome de poumon, en plus de son action sur c-Myc, le cramide bloque la fonction trans-activatrice de Sp1 et active le rpresseur Sp3 (protine de la famille de Sp1 se liant au mme site sur lADN mais ayant une fonction inhibitrice) avec pour effet linhibition de la transcription du gne hTERT (Wooten and Ogretmen 2005). De plus, il a t montr que la surexpression de p53 avait un effet rpresseur sur la transcription du gne hTERT, par inhibition de la liaison de Sp1 sur le promoteur (Kanaya, Kyo et al. 2000).

18

Comme la montr plusieurs reprises lquipe de Masaki Inoue, les estrognes activent la tlomrase dans des cellules de cancer du sein et de lendomtre. On retrouve deux sites de fixation du rcepteur aux estrognes (ERE : Estrogen Responsive Element) sur le promoteur de hTERT, lun -2677 pb, lautre coupl un site de reconnaissance de Sp1 -873 (Kyo, Takakura et al. 1999). Il a t montr que les estrognes pouvaient induire la transcription dhTERT soit directement par la fixation du rcepteur aux estrognes sur le promoteur, soit indirectement par lactivation dautres voies de signalisation (Kyo, Takakura et al. 1999; Wang, Kyo et al. 2002; Kimura, Ohmichi et al. 2004). La transcription du gne hTERT peut aussi tre induite par les facteurs de transcription de la famille Ets, qui sont retrouvs en aval de la voie des MAP Kinases (Mitogen Activated Protein Kinase) (Maida, Kyo et al. 2002; Goueli and Janknecht 2004). Le site consensus reconnu par Ets se situe en position -23 sur le promoteur de hTERT. NF- B semble aussi tre un rgulateur positif de la transcription du gne hTERT. Chez la souris en effet, NF- B rgule hTERT en se fixant un site consensus situ sur son promoteur (Yin, Hubbard et al. 2000). Chez lhomme, il a t montr que le virus HTLV-1 pouvait induire lexpression de hTERT via NF- B (Sinha-Datta, Horikawa et al. 2004). En ce qui concerne les rpresseurs du gne hTERT, en dehors de Mad1, on trouve WT-1 (Wilms Tumor 1) (Oh, Song et al. 1999), E2F-1 (Crowe, Nguyen et al. 2001; Lacerte, Korah et al. 2007), AP-1 (Takakura, Kyo et al. 2005) (mme si il peut aussi jouer un rle activateur (Alfonso-De Matte, Yang et al. 2002) et MZF-2 (Myeloid Zinc Finger 2) dont on a retrouv des sites de fixation sur le promoteur de hTERT (Fujimoto, Kyo et al. 2000).

1.2 Rgulation pigntique de la transcription Une rgulation pigntique du gne hTERT pourrait en partie expliquer la diffrence dexpression de la tlomrase entre les cellules normales et les cellules cancreuses. En effet, un traitement la trichostatine A, un inhibiteur dhistones dactylase, permet dobtenir lexpression de la tlomrase dans des cellules normales qui en sont initialement dpourvues (Cong and Bacchetti 2000; Takakura, Kyo et al. 2001). De plus, il semblerait que le facteur de transcription Sp1 joue un rle important dans cette ractivation de la transcription de hTERT, puisque une mutation sur les sites de fixation de 19

Sp1 sur le promoteur, bloque leffet de la trichostatine A (Takakura, Kyo et al. 2001). Plus rcemment, il a t montr quil existait bien un rarrangement de la chromatine permettant lactivation du gne hTERT lors de limmortalisation des cellules (Wang and Zhu 2003).

1.3 Rgulation par pissage alternatif Lexpression de la tlomrase peut tre rgule par un pissage alternatif de son ARN messager (Ulaner, Hu et al. 1998). Il existe en effet plusieurs variants dpissage rsultant de quatre insertions possibles de squences introniques et de deux dltions ( et ) (Figure 5), mais seulement 7 variants ont t retrouvs dans les cellules (Wick, Zubov et al. 1999; Yi, Shay et al. 2001). Linsertion 1 (38pb) et la dltion donc une protine tronque non fonctionnelle. La dltion (182pb) induisent un arrt prmatur de la traduction en amont du domaine transcriptase inverse, et donnent (36pb) a lieu au niveau du motif A du domaine transcriptase inverse, la protine obtenue, non tronque, nest pas fonctionnelle et agit comme dominant ngatif de la tlomrase active (Colgin, Wilkinson et al. 2000). Les autres insertions sont situes au niveau de la partie C-terminale et semblent aboutir des protines non fonctionnelles. Seul le transcrit entier, sans les dltions ni les insertions, donne lieu une protine hTERT active. Un changement au niveau du profil dpissage, donnant lieu une diffrence dexpression de la tlomrase, a t retrouv lors du dveloppement (Ulaner, Hu et al. 1998) ou entre les cellules normales et tumorales (Saeboe-Larssen, Fossberg et al. 2006).
Transcriptase inverse

Figure 5 : Organisation du gne hTERT avec les exons (en noir) et les sites dpissage alternatif (insertions 1 4 et dltions et ).

1.4 Rgulation post-traductionnelle

20

Plusieurs tudes ont montr que dans certaines cellules, malgr la prsence de lARN messager codant pour la forme active de hTERT, lactivit de la tlomrase restait indtectable (Rohde, Sattler et al. 2000; Liu, Hodes et al. 2001). Ceci montre que la protine obtenue aprs traduction du transcrit actif doit subir plusieurs tapes de modifications post-traductionnelles pour tre active : lassemblage des diffrentes sousunits, la translocation au noyau o sexerce son activit catalytique, et une phosphorylation qui va permettre daugmenter son activit. Il semblerait en fait que la sous-unit catalytique soit dabord transloque au noyau o a lieu lassemblage avec la sous-unit hTR (pour revue, (Aisner, Wright et al. 2002)) (Figure 6).

Cytoplasme

Noyau
Hsp70

WT

B
WT
14-3-3

Hsp90

p23

Forme sauvage et variants dpissage de hTERT

WT P WT P

C
hTR
H/ACA

PKC Akt

Figure 6 : Rgulation post-traductionnelle de la tlomrase. A, translocation au noyau et squestration grce la protine 14-3-3 ; B, assemblage grce aux protines chaperonnes Hsp90 et Hsp70 ; C, phosphorylation par PKC et Akt (daprs (Aisner, Wright et al. 2002)).

1.4.1.

Translocation au noyau

Aprs sa synthse, la sous-unit catalytique hTERT doit tre transloque au noyau afin de permettre son assemblage avec la sous-unit ribonuclique, puis daccder aux tlomres. Les protines impliques dans cette translocation ne sont pas compltement connues, mais il semblerait que la protine 14-3-3 y participe en sassociant hTERT, ce qui bloque son export vers le cytoplasme. En effet, lexpression dun mutant dominant-ngatif de 14-3-3 dans les cellules induit une

21

redistribution de la tlomrase vers le cytoplasme (Seimiya, Sawada et al. 2000), et de plus, des souris dficientes pour la protine 14-3-3 ont des tlomres raccourcis avec une augmentation du nombre de fusions tlomre-tlomre (Dhar, Squire et al. 2000). Plus rcemment, il a t montr que le stress oxydatif pouvait induire une relocalisation cytoplasmique de la sous-unit catalytique hTERT, et cela grce une phosphorylation sur la tyrosine 707 par une kinase de la famille Src (Haendeler, Hoffmann et al. 2003). 1.4.2. Repliement et assemblage

Comme nous lavons vu prcdemment, pour tre active, la tlomrase doit soligomriser. Chaque oligomre contient en gnral deux sous-units catalytiques hTERT et deux sous-units ribonucliques hTR, ce qui suggre lexistence de deux sites catalytiques par enzyme (Wenz, Enenkel et al. 2001). Le variant dpissage de la protine hTERT est capable de sassembler avec la sous-unit hTR mais lenzyme obtenue nest pas fonctionnelle et agit comme dominant ngatif de la tlomrase (Colgin, Wilkinson et al. 2000). Loligomrisation a lieu dans le noyau, sous le contrle de protines chaperonnes Hsp90, Hsp70 et p23, qui sont capables de se fixer hTERT, comme cela a t montr dans le cancer de la prostate (Akalin, Elmore et al. 2001). Hsp70 se fixe hTERT en labsence de hTR et le complexe se spare lorsque hTERT se retrouve dans une conformation active, ce qui suggre que cette chaperonne joue un rle dans le repliement de la protine (Forsythe, Jarvis et al. 2001). Hsp90 et p23, quant elles, jouent un rle dans lassemblage de hTERT avec hTR et peuvent rester fixes avec le complexe tlomrase (Forsythe, Jarvis et al. 2001). Plus rcemment, lquipe de Jarstfer a montr que lon pouvait inhiber la tlomrase dans des cellules en empchant loligomrisation de la protine grce des oligonuclotides capables de reconnatre et de se fixer hTR sur les domaines P1/P3 et CR4-CR5 (Keppler and Jarstfer 2004). 1.4.3. Phosphorylation

Il a t montr que lactivation du lymphocyte T tait associe une activation de la tlomrase, or dans ce cas, il ny a pas daugmentation de la

22

quantit de protine hTERT dans la cellule. Cette augmentation dactivit est uniquement due une relocalisation cellulaire et une phosphorylation de la tlomrase (Liu, Hodes et al. 2001). En effet, de nombreuses tudes ont montr que hTERT pouvait tre phosphoryle ou dphosphoryle par certaines kinases et phosphatases, avec pour effet une modulation de son activit. Cest le cas de plusieurs srine/thronine kinases de la famille des PKCs qui induisent une activation de la tlomrase (Ku, Cheng et al. 1997; Li, Zhao et al. 1998; Yu, Lo et al. 2001). De mme, la protine kinase B (Akt) implique dans la voie de la PI3kinase, phosphoryle hTERT sur le rsidu srine 824 ce qui a pour effet daugmenter son activit (Kang, Kwon et al. 1999). La phosphatase PP2A a au contraire un effet inhibiteur de la tlomrase en dphosphorylant les sous-units hTERT prsentes dans le noyau (Li, Zhao et al. 1997). c-Abl, quant lui, peut phosphoryler hTERT en rponse des dommages de lADN causes par des radiations ionisantes, mais dans ce cas, cette phosphorylation aura un effet inhibiteur sur lactivit de la tlomrase (Kharbanda, Kumar et al. 2000). Enfin, comme nous lavons vu prcdemment, hTERT peut tre phosphoryle sur le rsidu tyrosine 707 par une kinase de la famille Src, ce qui induit sa relocalisation cytoplasmique (Haendeler, Hoffmann et al. 2003).

2. Rgulation de la sous-unit ribonuclique

Comme nous lavons vu plus haut, la sous-unit ribonuclique hTR, nest pas le facteur limitant de lactivit tlomrase et est exprime de faon ubiquitaire. On retrouve cependant de rares cas o la quantit de hTR dans la cellule peut varier. 2.1 Rgulation transcriptionnelle et post-transcriptionnelle Le promoteur de hTR est long de 231pb et contient des lots CpG susceptibles dtre mthyls, ce qui est le cas dans certaines lignes tlomrase ngatives o le maintien de la longueur des tlomres a lieu grce au mcanisme ALT (Alternative Lenghtening of Telomeres) (Hoare, Bryce et al. 2001). Dautre part, le niveau dexpression de hTR peut

23

tre augment dans certaines cellules tumorales qui surexpriment hTERT (Yi, Tesmer et al. 1999).

2.2 Stabilit de hTR Plusieurs protines pouvant sassocier hTR jouent un rle dans sa stabilisation. Cest le cas de protines H/ACA telles que la dyskrine chez lhomme, qui se fixe au domaine H/ACA de la sous-unit hTR et permet sa stabilisation et donc son accumulation dans la cellule (Mitchell, Cheng et al. 1999; Cohen, Graham et al. 2007). De plus, la dyskrine est mute dans la diskratose congnitale, maladie au cours de laquelle on retrouve une inhibition de la tlomrase et un raccourcissement des tlomres (Vulliamy, Marrone et al. 2001). Dautres protines telles que les hnRNP C1 et C2 (Ford, Suh et al. 2000), La (Ford, Shay et al. 2001), et L22 (Le, Sternglanz et al. 2000; Bachand, Triki et al. 2001) sont associes hTR et sont susceptibles de jouer un rle dans la stabilisation de hTR, sa localisation cellulaire ou lassemblage de la tlomrase.

3. Les signaux intra et extracellulaires conduisant la rgulation de la tlomrase

3.1 Les hormones Comme nous lavons vu prcdemment, la tlomrase peut tre rgule par certaines hormones. En effet, les estrognes ont un effet activateur de hTERT en agissant directement sur le promoteur via le rcepteur aux estrognes (Kyo, Takakura et al. 1999), mais aussi indirectement en activant la voie PI3-kinase/Akt (Kimura, Ohmichi et al. 2004). Cet effet activateur a t retrouv dans plusieurs types cellulaires dont le cancer du sein (Kyo, Takakura et al. 1999), les cellules pithliales ovariennes (Misiti, Nanni et al. 2000), le cancer de lendomtre (Boggess, Zhou et al. 2006), le cancer de la prostate (Nanni, Narducci et al. 2002). En revanche, leffet du tamoxifne, connu comme modulateur slectif du rcepteur aux estrognes (SERM), nest pas aussi marqu, puisquil inhibe lactivit de la tlomrase dans les cellules de cancer du sein, mais linduit dans le cancer

24

de lendomtre (Wang, Kyo et al. 2002). Dautre part, la progestrone dont le rle est antagoniste aux estrognes, a un effet inhibiteur sur la tlomrase dans les cellules de cancer du sein, travers une induction de la voie des MAP kinases et de lexpression de p21 (Wang, Kyo et al. 2000). Enfin, les andrognes peuvent rguler la tlomrase de faon diffrente entre les tissus sains et les tissus cancreux. En effet, la castration chez le rat ou chez le singe induit une augmentation de lactivit tlomrase dans les cellules de prostate saine (Meeker, Sommerfeld et al. 1996; Ravindranath, Ioffe et al. 2001), alors que le niveau de tlomrase augmente dans les cellules de cancer de la prostate lorsque lon enlve les andrognes (Iczkowski, Huang et al. 2004). 3.2 Les facteurs de croissance Les facteurs de croissance jouent un rle essentiel dans la rgulation positive ou ngative de la prolifration et de la diffrenciation dans de nombreux types cellulaires. Par ailleurs, il savre que plusieurs facteurs de croissance ont t impliqus dans la rgulation de la tlomrase. LEGF (epidermal growth factor) par exemple, qui induit la transcription du gne de la tlomrase travers la voie Ras/MEK/ERK et les facteurs de transcription Ets (Maida, Kyo et al. 2002; Budiyanto, Bito et al. 2003). De mme, le FGF (fibroblast growth factor) (mais pas le VEGF (vascular endothelial growth factor)) a un effet activateur de la transcription du gne hTERT (Kurz, Hong et al. 2003). LIGF-1 (Insulinlike growth factor-1) active de mme la transcription du gne hTERT dans les cellules de cancer de la prostate, par la voie PI3K/Akt (Wetterau, Francis et al. 2003). En revanche, le TGF- (transforming growth factor- ) a un effet inhibiteur sur la transcription du gne hTERT dans de nombreux types cellulaires. Cette inhibition peut tre mdie par diffrents facteurs de transcription, dont notamment par le couple c-myc/Mad-1 (Yang, Kyo et al. 2001), SIP-1 (Lin and Elledge 2003) ou directement par Smad3, lun des rgulateurs principaux de la voie du TGF- (Li, Xu et al. 2006). 3.3 Les cytokines Comme lont montr Engelhardt et son quipe, la tlomrase peut tre rgule par des cytokines de lhmatopose. En effet, lorsque des cellules souches hmatopotiques sont soumises un mlange de cytokines permettant leur prolifration (SCF (Stem cell

25

factor), IL3, IL6 (Interleukine 3 et 6), G-CSF (Granulocyte-colony stimulating factor) et Epo (Erythropotine)), on observe une augmentation de lactivit tlomrase pendant les 7 premiers jours de culture (Engelhardt, Kumar et al. 1997). De plus, dans des cellules de mylome multiple, lIL6 et lIGF-1 stimulent lactivit de la tlomrase via la voie PI3K/Akt/NF- B, et ce sans interfrer sur la transcription du gne hTERT (Akiyama, Hideshima et al. 2002). Par ailleurs, plusieurs tudes montrent que certaines cytokines inhibitrices ou ligands de mort pouvaient avoir de mme un effet sur la tlomrase. Ainsi, lintrferon- , (IFN- ) rprime la transcription du gne hTERT (Xu, Erickson et al. 2000), de mme pour lintrferon- (IFN- ) dont leffet est mdi par lIRF-1 (interferon regulatory factor-1) (Lee, Kim et al. 2003; Lee, Kim et al. 2005). Au sein de notre quipe nous avons montr que le TNF(tumor necrosis factor- ) inhibait lactivit de la tlomrase dans les cellules hmatopotiques de la ligne mylode, ce qui conduit un tat de snescence (Beyne-Rauzy, Recher et al. 2004). Cependant des travaux antrieurs montraient que le TNF- avait, au contraire, un effet activateur indirect de la tlomrase dans des cellules de mylome multiple (Akiyama, Hideshima et al. 2003) ou dans les lymphocytes (Akiyama, Yamada et al. 2004), en permettant sa translocation au noyau grce linteraction avec NF-kB p65.

26

IV. Rles de la tlomrase

1. Maintien de la longueur des tlomres
La fonction primordiale de la tlomrase est le maintien de la longueur des tlomres par ajout leur extrmit de squences rptes TTAGGG. En effet, dans des cellules dpourvues de tlomrase, on assiste un raccourcissement des tlomres au cours des divisions cellulaires. Lorsque les tlomres atteignent une taille critique, la cellule entre en snescence rplicative, ce qui a pour effet de stopper sa prolifration et finalement daboutir la mort de la cellule. On comprend ainsi que les cellules ayant une importante capacit prolifrative, telles que les cellules prognitrices des tissus renouvellement rapide (tissus cutans, intestinaux et hmatopotiques), prsentent une forte activit tlomrase, alors quelle est trs faiblement exprime au sein des autres cellules somatiques, qui nont pas la ncessit de raliser un nombre lev de divisions cellulaires. Lactivation de la tlomrase est le mcanisme de maintien de la longueur des tlomres le plus frquent dans les cellules tumorales humaines (85 90 % des cas) (Kim, Piatyszek et al. 1994), avec le plus souvent une surexpression de la sous-unit catalytique de lenzyme. Cependant dans certains cas particuliers, on peut trouver des cellules immortelles ou tumorales sans activit tlomrase dtectable (Bryan, Englezou et al. 1995; Bryan, Englezou et al. 1997). Dans ce cas, il existe un mcanisme dallongement alternatif des tlomres, appel ALT (Alternate Lenghtening of Telomeres), mettant en jeu un systme de recombinaisons. Des tudes ont en effet montr que llongation des tlomres a lieu, dans ce cas, par hybridation dun brin dADN dun tlomres sur le brin complmentaire dun autre tlomre (Dunham, Neumann et al. 2000; Reddel 2003), avec pour consquence directe une trs grande htrognit de longueurs des tlomres. 1.1 La snescence rplicative Comme nous lavons vu plus haut, lorsque aucun mcanisme de maintien des tlomres nest mis en jeu, les extrmits tlomriques raccourcissent chaque rplication de lADN. Lorsque les tlomres atteignent une taille critique, les cellules arrtent de prolifrer (blocage en G1 du cycle cellulaire), et acquirent une morphologie particulire (largissement et aplatissement des cellules) avec des fonctions altres et une induction de

27

la SA- -galactosidase, qui caractrisent un tat de snescence rplicative (Hayflick 1965; Dimri, Lee et al. 1995; Campisi, Kim et al. 2001). Cet arrt en G1 est induit par la rpression de gnes de progression dans le cycle cellulaire, et par la rgulation positive de gnes inhibiteurs tels que p21 et p16. Il est important de noter que la prsence des suppresseurs de tumeurs pRB et p53 est ncessaire lentre en snescence, puisque linduction de p16INK4a bloque le cycle cellulaire en inhibant la phosphorylation de pRB, et que lactivation du facteur de transcription p53 aboutit la rgulation positive de p21WAF1 (Gire and Wynford-Thomas 1998). Lexpression ectopique de la tlomrase permet dviter lentre en snescence grce la prvention du raccourcissement tlomrique, et donc limmortalisation de lignes. Ce processus de snescence devrait permettre priori de protger lorganisme de lapparition de cancer en empchant les cellules de raliser un trop grand nombre de divisions cellulaires (Campisi, Kim et al. 2001). Cependant des tudes montrent que ltat de snescence saccompagne dune forte instabilit gntique. En effet, le raccourcissement des tlomres induit une augmentation du nombre de recombinaisons non homologues avec fusions chromosomiques et fusions de chromatides surs (Hande, Samper et al. 1999), ce qui pourrait au contraire favoriser lmergence de certains oncognes. 1.2 La snescence prmature induite par le stress Ds 1998, plusieurs travaux rapportent que certains agents anti-tumoraux (inhibiteurs de topoisomrase II, actinomycine D, blomycine) utiliss des doses subltales sont susceptibles dinduire un phnotype de snescence dans des fibroblastes humains en culture, plaant ainsi la snescence comme une nouvelle modalit de rponse cellulaire au stress (Michishita, Nakabayashi et al. 1998; Robles and Adami 1998). Ces observations ont permis dintroduire le concept de snescence prmature induite par le stress (SPIS ou SIPS pour stress-induced premature senescence). Les cellules entrant en SPIS prsentent les mmes caractristiques que les cellules en snescence rplicative, avec arrt en phase G1 du cycle cellulaire, raccourcissement des tlomres, modifications morphologiques et expression de SA- -galactosidase. De mme, il sagit dun processus hautement rgul, impliquant des lments de contrle tels que la tlomrase, p53, p16 ou p21 (pour revue voir (Mansat-De Mas 2003)), et il saccompagne de la mme faon dune forte instabilit chromosomique. Cependant, contrairement la snescence rplicative, les 28

causes directes de la SPIS ne sont pas entirement connues. En effet, la plupart des tudes montrent quil nexiste pas de corrlation entre la SPIS et lactivit de la tlomrase (Wang, Wong et al. 1998; Ren, Xia et al. 2001; Gorbunova, Seluanov et al. 2002). Lentre en SPIS serait plus probablement lie une dsorganisation des rgions tlomriques due aux dommages de lADN gnrs par les agents anti-tumoraux, et une rgulation des protines p53, p21 et p16.

2. Prolifration
Comme nous lavons vu prcdemment, il est acquis que la tlomrase joue un rle indirect dans le maintien des capacits prolifratives des cellules long terme. En effet, labsence de tlomrase induit le raccourcissement des tlomres et donc lentre en snescence rplicative, ce qui provoque un arrt de la prolifration par blocage du cycle cellulaire. Cependant, limplication directe de la tlomrase dans la rponse prolifrative a t trs peu documente. Il a tout dabord t montr lexistence dune corrlation entre les phases du cycle et lactivit tlomrase, avec une augmentation du niveau de tlomrase dans la cellule lors de la phase S, mais sans que le lien direct nai t dtermin (Zhu, Kumar et al. 1996). Plus rcemment, quelques travaux montrent un lien direct entre lexpression de la tlomrase et la croissance cellulaire. En effet, la surexpression de hTERT dans des cellules fibroblastiques permet lacquisition dun avantage prolifratif court et long terme, c' est--dire un temps de doublement de population plus rapide (Forsythe, Elmore et al. 2002). Au contraire, lorsque lexpression de la tlomrase est inhibe par des techniques de siRNA, on assiste ds trois jours une diminution de la prolifration avec augmentation du temps de doublement de population, alors quon ne note pas encore de raccourcissement des tlomres (Gandellini, Folini et al. 2007). Par ailleurs, Li et coll. ont montr que linhibition de hTERT induisait des changements dexpressions de nombreux gnes dont certains impliqus dans la prolifration (Li, Crothers et al. 2005).

3. Protection contre lapoptose

Outre son rle connu dans le maintien des capacits prolifratives long terme des cellules, ces dernires annes, dautres rles ont t attribus la tlomrase, et notamment

29

pour une trs large part, dans la protection contre lapoptose. En effet, de nombreux travaux montrent par inhibition ou surexpression de la tlomrase, que cette enzyme intervient dans la protection contre divers stress, tels que le sevrage en srum ou cytokine, lexposition aux agents genotoxiques, aux radiations ionisantes, aux inhibiteurs dhistone deactylases, ou aux ligands de mort (Akiyama, Yamada et al. 2002; Misawa, Tauchi et al. 2002; Wu, Meng et al. 2005; Li, Ferguson et al. 2006; Gao and Chen 2007). Cependant, le ou les mcanismes par lesquels la tlomrase agit contre linduction de lapoptose restent largement discuts. Lune des premires possibilits est que hTERT puisse intervenir en bloquant lapoptose induite par p53 comme montr dans des cellules de lymphomes de Burkitt (Rahman, Latonen et al. 2005) ou de cancer du sein (Cao, Li et al. 2002). Il semblerait de plus que dans les deux cas, cet effet soit indpendant de lactivit catalytique de la tlomrase, puisque un mutant dominant ngatif de hTERT catalytiquement inactif a le mme effet de protection que le sauvage. En revanche, dautres travaux montrent au contraire, que lactivit catalytique de hTERT et sa liaison la protine 14-3-3 sont indispensables sa fonction anti-apoptotique (Zhang, Chan et al. 2003). Dans ce cas, il semblerait que hTERT agisse un stade pr-mitochondrial. Ce dernier point est confirm par une rcente tude dans laquelle il apparat que linhibition de hTERT par des siRNA sensibilise les cellules la mort induite par des agents genotoxiques, en agissant au niveau de la mitochondrie par activation de Bax (Massard, Zermati et al. 2006). De mme, la tlomrase protge les neurones de la mort induite par le NMDA en modifiant les flux calciques au niveau de la mitochondrie, via une voie indpendante de son rle au niveau des tlomres (Kang, Choi et al. 2004). Enfin, dans certains contextes, la tlomrase peut agir indirectement sur la transcription de certains gnes impliqus dans la rponse apoptotique. Ainsi la surexpression de hTERT dans des cellules pithliales mammaires inhibe des gnes pro-apoptotiques tels que TRAIL (Smith, Coller et al. 2003).

4. Rparation de lADN
Plusieurs tudes ont pu observer que la tlomrase permettait de protger les cellules contre les dommages lADN induits par les rayonnements ionisants (Pirzio, Freulet-Marriere et al. 2004; Nakamura, Masutomi et al. 2005; Serakinci, Christensen et al. 2007), ou tout simplement de stabiliser le gnome (Sharma, Gupta et al. 2003). En plus de son rle dlongation des tlomres, il semblerait que la tlomrase par sa simple

30

localisation au niveau des tlomres, puisse protger les extrmits chromosomiques en empchant la reconnaissance par les systmes de rparation des cassures doubles brins (Chan and Blackburn 2002). La tlomrase entrerait donc dans la composition dune coiffe tlomrique protectrice. Dautre part, la tlomrase semble tre directement implique dans les processus de rparation de lADN. En effet, Masutomi et coll. ont pu montrer que linhibition de la tlomrase tait lorigine dune radiosensibilit accrue due une diminution de la rparation des cassures doubles brins, comme le montre la diminution de la phosphorylation de -H2AX (Masutomi, Possemato et al. 2005). De mme, hTERT serait implique dans le systme NER (nucleotide excision repair), dans la rparation des dommages causs par les UV et dans le systme NHEJ (non homologous end-joining) (Shin, Kang et al. 2004). Cependant, les mcanismes par lesquels hTERT agit sur les voies de rparation de lADN ne sont pas encore totalement identifis. Certaines tudes suggrent que la tlomrase puisse sassocier et interagir avec des protines appartenants aux cascades de rparation des dommages de lADN, comme par exemple la protine Ku70/80 (Chai, Ford et al. 2002). Dautre part, Sharma et son quipe ont mis en vidence une augmentation de lexpression de nombreuses protines intervenant dans les voies de rparation de lADN, dont MSH6 et MLH1 pour le MMR (mismatch repair), XRCC4 et 9 pour le NHEJ ou Rad51 pour la recombinaison homologue (Sharma, Gupta et al. 2003).

5. Diffrenciation
La tlomrase semble de plus jouer un rle dans le processus de diffrenciation. En effet, mme si son effet reste encore controvers, des tudes montrent que la modulation de lexpression de la tlomrase intervient dans la capacit de diffrenciation terminale des cellules. Divers travaux montrent que lors de la diffrenciation des cellules de la ligne HL60 en mgacaryocytes induite par le DMSO, lATRA (all-trans-retinoic acid), ou le TPA, on assiste une diminution de lactivit de la tlomrase due linhibition de la transcription du gne hTERT (Xu, Gruber et al. 1999; Koyanagi, Kobayashi et al. 2000; Wu, Chen et al. 2002), et que cette inhibition prcde lentre en diffrenciation (Liu, Berletch et al. 2004). De nombreuses tudes plus rcentes sur ce sujet ont permis de dcouvrir les mcanismes de cette inhibition. Ainsi, il semblerait que lors du traitement par le DMSO, on assiste un switch du complexe Myc/Max vers le complexe Mad1/Max sur le promoteur du gne hTERT, ainsi qu une diminution de lactylation des histones en

31

dbut de diffrenciation qui se poursuit par une mthylation au niveau du promoteur dhTERT lors de la diffrenciation terminale, permettant dteindre compltement le gne (Xu, Popov et al. 2001; Lopatina, Poole et al. 2003). Dautres modles permettent galement dobserver une diminution de lexpression de la tlomrase lors de la diffrenciation des cellules trophoblastiques induite par le TGF- (Rama, Suresh et al. 2001), des cellules germinales des testicules (Schrader, Burger et al. 2002), des cellules pithliales (Crowe, Nguyen et al. 2005), ou des cellules souches embryonnaires de souris (Wang, Hu et al. 2007). Cependant, le rle rel de la tlomrase dans ce processus reste largement controvers. Ainsi, certains travaux ont montr que la surexpression de hTERT avait pour effet de bloquer la diffrenciation des cellules neuronales (Richardson, Nguyen et al. 2007), des cellules pithliales bovines (Wang, Feng et al. 2005) ou des kratinocytes humains (Cerezo, Stark et al. 2003). En revanche, dans certains modles, la surexpression dhTERT na pas deffet sur la diffrenciation (Robertson, Li et al. 2005) ou au contraire un effet activateur comme dans les cellules hmatopotiques (Zimmermann, Glaser et al. 2004; Armstrong, Saretzki et al. 2005) ou dans les cellules stromales du tissu adipeux (Jun, Lee et al. 2004). Daprs ces rsultats, limplication de la tlomrase dans le processus de diffrenciation cellulaire semble tre dpendant du type de tissus, et des tudes plus pousses permettront de dterminer plus prcisment son mcanisme daction.

6. Modles murins de modulation de la tlomrase

La biologie des tlomres et de la tlomrase chez la souris est trs diffrente de celle de lhomme. Les tlomres sont en effet 4 10 fois plus longs chez la souris de laboratoire Mus musculus, que chez lhomme (Blasco, Lee et al. 1997; Zijlmans, Martens et al. 1997), et on retrouve une activit de la tlomrase dans la plupart des cellules somatiques (Prowse and Greider 1995). Ainsi, la snescence rplicative telle quon la connat chez lhomme, nest pas retrouve lors du processus de vieillissement chez la souris. Les souris knockout (KO) pour la tlomrase sont obtenues en inhibant le gne de la sous-unit ribonuclique mTERC, par recombinaison homologue (Blasco, Lee et al. 1997). Les souris mTERC-/- obtenues sont viables, fertiles et ne prsentent pas danomalies morphologiques. La mesure de la longueur des tlomres indique, comme attendu, une perte denviron 120pb par division cellulaire, mais tant donn les trs longs tlomres chez la souris, il faut plusieurs gnrations de souris avant davoir un raccourcissement

32

significatif. En effet, les croisements entre souris mTERC-/- permettent dobtenir des effets physiologiques partir de la quatrime gnration. Ces effets sont observs au niveau de tissus forte prolifration, dont notamment, lhmatopose et le systme reproductif. Pour ce qui est de lhmatopose, on observe une diminution du nombre de prcurseurs, une inhibition de la prolifration des lymphocytes T et B sous stimulation et une rduction de la fonction du centre germinatif (Lee, Blasco et al. 1998; Herrera, Martinez et al. 2000). Une diminution de la fcondit est dabord observe partir de la quatrime gnration, pour aboutir une strilit la sixime gnration, avec une atrophie des testicules chez les mles et une diminution du nombre dovocytes produits par ovulation chez les femelles (Lee, Blasco et al. 1998). De faon intressante, la rexpression de mTERC dans les souris dficientes, mme pour les gnrations tardives, permet de rallonger les tlomres et ainsi de retrouver une physiologie normale (Samper, Flores et al. 2001). Tous ces rsultats montrent que le renouvellement long terme des tissus forte prolifration est dpendant de lintgrit des tlomres et de lactivit de la tlomrase. Dautre part, des travaux ont montr que les souris mTERC-/- aux gnrations tardives avaient une frquence plus leve de tumeurs spontanes (lymphomes ou tratocarcinomes) en comparaison avec des souris normales ou mTERC-/- aux gnrations prcoces (Rudolph, Chang et al. 1999), et que linvalidation de p53 en mme temps que la tlomrase conduit une aggravation de linduction de ces tumeurs (Artandi and DePinho 2000). Ceci peut tre expliqu par une augmentation de linstabilit chromosomique avec apparition daneuplodie et de fusions chromosomiques bout bout. Ces rsultats confirment limplication de la tlomrase dans la carcinogense.

7. Tlomrase et cancer
Comme nous lavons vu prcdemment, except les cellules fort renouvellement, les cellules somatiques prsentent une activit tlomrase quasiment nulle et ne possdent donc pas de mcanisme de maintien de la longueur des tlomres. En revanche, toutes les cellules cancreuses possdent un moyen dviter lrosion tlomrique, soit par le mcanisme dALT, soit grce la prsence dune activit tlomrase leve (85 90 % des tumeurs). Lexpression ectopique de la tlomrase permet en effet dimmortaliser les cellules somatiques normalement voues un nombre restreint de divisions cellulaires (Bodnar, Ouellette et al. 1998; Vaziri and Benchimol 1998), voire de les transformer lors

33

dune coexpression avec Ras ou lantigne T du virus SV40. Ceci suggre un lien entre tlomrase et cancer, et soulve la question de la dualit de cette enzyme vis--vis du processus de tumorignse. En effet, le statut de la tlomrase ainsi que la longueur des tlomres, peuvent apparatre la fois comme un processus anti et pro-tumoral.

Longueur tlomres Cellules transduites par hTERT : immortalisation

Snescence
p53, Rb

Apoptose
hTERT/ALT oncognes

Cellules tumorales

Crise
Instabilit gntique

Apoptose
Divisions cellulaires

Figure 7 : Implication des tlomres et de la tlomrase dans la tumorignse.

Des tudes montrent que lactivit tlomrase est trs faible dans les lsions prtumorales et quelle augmente lors de la progression tumorale (Engelhardt, Drullinsky et al. 1997; Kolquist, Ellisen et al. 1998; Yan, Coindre et al. 1999). De plus, dans la plupart des cellules cancreuses, on observe la prsence dune activit tlomrase leve avec des tlomres courts et de frquentes anomalies cytogntiques impliquant les tlomres (Gisselsson, Jonson et al. 2001; Rudolph, Millard et al. 2001). Ces donnes exprimentales permettent de proposer un modle expliquant la dualit de la tlomrase dans le processus de tumorignse. Lorsque les cellules somatiques, prsentant une activit tlomrase nulle ou trs faible, ralisent un grand nombre de divisions cellulaires, on observe un raccourcissement des tlomres, jusqu lentre en phase de snescence, qui saccompagne dun arrt net de la prolifration. A ce stade, il existe un point de contrle impliquant p53 et la protine Rb qui empche les cellules de repartir en prolifration. Ainsi, la snescence peut apparatre comme un mcanisme anti-tumoral qui bloque la

34

prolifration des cellules ayant ralis un trop grand nombre de divisions. En revanche, si malgr tout, une anomalie cellulaire permet la cellule de passer ce point de contrle et de continuer prolifrer, le raccourcissement des tlomres saccentue et la cellule entre en phase de crise qui saccompagne dune trs forte instabilit gntique. Lors de cette phase, un grand nombre de cellules entrent en apoptose, mais une partie dentre elles acquirent des mutations permettant lexpression doncognes, et le rtablissement dun mcanisme de maintien de la longueur des tlomres avec une surexpression de la tlomrase, ce qui peut aboutir la naissance dun clone cancreux. Dans ce cas, la tlomrase peut apparatre comme une protine pro-carcinogne (Figure 7). Son rle dans linduction de la prolifration et la protection contre lapoptose peut aussi participer ce rle carcinogne. Cependant, on sait que si les cellules expriment la tlomrase avant le raccourcissement tlomrique, celle-ci permet de protger les extrmits des chromosomes et donc dviter lapparition dinstabilit gntique. En effet, il a t montr que les modles de souris invalides pour la tlomrase (souris mTERC-/-), prsentaient un plus grand nombre de tumeurs spontanes dues linstabilit gntique provoque par labsence de tlomrase fonctionnelle (Rudolph, Chang et al. 1999). Ceci porterait croire que la tlomrase nest pas indispensable au processus de tumorignse et que le maintien des tlomres peut tre ralis par dautres mcanismes tels que lALT.

35

V. Tlomrase et hmatopose
1. Lhmatopose
Lhmatopose est le processus qui permet dobtenir lensemble des cellules sanguines partir des cellules souches pluripotentes de la moelle osseuse. A partir dun petit contingent de cellules souches hmatopotiques (CSH), des tapes successives de prolifration et de diffrenciation vont permettre de produire jusqu 1012 cellules sanguines matures par jour chez lhomme. Lhmatopose est compose de huit lignes, dont deux sont lymphodes, permettant dobtenir les lymphocytes B et T, et six mylodes, permettant dobtenir les plaquettes (ligne mgacaryocytaire), les hmaties (ligne rythrode), les monocytes (ligne monocytaire), et les polynuclaires neutrophiles, osinophiles ou basophiles (lignes granulocytaires) (Figure 8). Lhmatopose peut tre divise en trois compartiments virtuels (figure 8): le compartiment des CSH qui contient des cellules pluripotentes longue dure de vie, qui lors dune stimulation sont capables de sortir de quiescence, de sautorenouveler et de se diffrencier vers toutes les lignes sanguines. Ceux sont ces CSH qui permettent de reconstituer lhmatopose lors de greffe de moelle. le compartiment des progniteurs est constitu par la descendance des CSH. Ces cellules sont capables de prolifrer et de se diffrencier sous laction de facteurs de croissance, mais ont perdu leur capacit dauto-renouvellement. Les progniteurs les plus primitifs sont capables de se diffrencier dans plusieurs lignes diffrentes (le progniteur mylode ou lymphode), et leur distinction des CSH nest pas nette puisquils peuvent permettre de reconstituer une hmatopose in vivo long terme. Puis, au fur et mesure de la diffrenciation, les progniteurs finissent par devenir monopotents en sengageant dans un type de ligne spcifique (par exemple, le CFU-M, progniteur monocytaire). le compartiment des prcurseurs est compos des cellules plus matures, morphologiquement reconnaissables ayant acquis des proprits spcifiques de

36

chaque ligne. A ce stade, les cellules ne ralisent plus que quelques divisions cellulaires. Contrairement aux autres compartiments, ceux sont les seules cellules que lon peut dtecter dans la moelle osseuse. A la fin de la maturation les cellules hmatopotiques terminales passent dans la circulation sanguine.

Figure 8 : Schma reprsentatif de lhmatopose normale permettant dobtenir les cellules matures prsentes dans le sang partir des cellules souches de la moelle osseuse. CFU-MK : colony forming unitmegakaryocitic, BFU-E : burst forming unit-erythroid, CFU-E : colony forming unit-erythroid, CFU-GM : colony forming unit-granulo/mono, CFU-M : colony forming unit-monocytic, CFU-G : colony forming unitgranulocitic, CFU-Eo : colony forming unit-eosinophile, CFU-Bas : colony forming unit-basophile, PNN : polynuclaire neutrophile, PNEo : polynuclaire osinophile, natural killer. PNBaso : polynuclaire basophile, NK :

Chaque cellule porte sa surface des marqueurs phnotypiques permettant de les diffrencier et de connatre leur stade de maturation et la ligne dans laquelle la cellule est engage. Ainsi, lexpression du marqueur CD33 indiquera une diffrenciation granulocytaire, alors quune cellule sengageant dans la ligne rythrode exprimera le marqueur CD36 puis le CD235a (glycophorine A) la fin de la maturation. Lexpression 37

des diffrents marqueurs correspondant chaque ligne sont rsums dans le tableau 1. Le phnotype des cellules souches hmatopotiques reste flou. Il est cependant admis que les CSH ou les progniteurs prcoces nexpriment que trs peu de marqueurs, et que lon peut les isoler sur la base de lexpression du marqueur CD34 et labsence dautres marqueurs. Ainsi, des tudes montrent que les CSH se trouvent dans les fractions cellulaires CD34+/CD38neg (Conneally, Cashman et al. 1997), ou CD34+/Thy-1+/Lin- (Sutherland, Hogge et al. 1996), ou CD34+/CD45RAfaible (Lansdorp, Sutherland et al. 1990). Lhmatopose est un mcanisme hautement rgul par les facteurs de croissance et les inhibiteurs endognes. Ainsi, pour se diffrencier vers une ligne spcifique, les cellules doivent tre en contact avec un mlange complexe de cytokines. De nombreuses tudes ont permis didentifier les mlanges de cytokines correspondant chaque ligne hmatopotique, permettant ainsi dobtenir in vitro les cellules matures de chaque ligne partir de culture liquide de CSH. Les principales cytokines correspondant chaque ligne sont rsumes dans le tableau 1. Ces cytokines sont produites par les cellules du stroma mdullaire, les lymphocytes, ou libres dans la circulation systmique comme lrythropotine (Epo) synthtise par le rein. Certains facteurs vont au contraire avoir un effet dltre sur lhmatopose dont le TNF (Tumor necrosis factor ), lIFN (interfron ), lIL-1 (interleukine 1) ou le TGF (transforming growth factor ).

Tableau 1 : Marqueurs membranaires et cytokines caractristiques de chaque ligne hmatopotique.

Ligne hmatopotique Mgacaryocytaire Erythrode Monocytaire Granulocytaire neutrophile Granulocytaire basophile Granulocytaire osinophile Lymphode B Lymphode T

Marqueurs CD51, CD61, CD41a, CD41b CD36, CD71, GPA CD14, CD11b CD16 CRTH2 CRTH2 CD19 CD2

Cytokines IL6, Thrombopotine SCF, IL3, rythropotine GM-CSF, M-CSF GM-CSF, G-CSF, IL4 GM-CSF, IL4 GM-CSF, IL5 SCF, IL3 SCF, IL7

38

2. La tlomrase et les tlomres dans les cellules hmatopotiques

2.1. Tlomrase et diffrenciation hmatopotique Comme nous lavons vu prcdemment, contrairement la plupart des cellules somatiques, les cellules hmatopotiques qui prsentent un fort taux de prolifration ont une activit tlomrase dtectable. Des tudes plus pousses montrent que les cellules souches primitives du tissu hmatopotique quiescentes ont une activit tlomrase quasiment nulle, et quon observe une importante activation de cette enzyme dans les progniteurs prcoces (Chiu, Dragowska et al. 1996), c' est--dire lorsque les cellules sont stimules par des facteurs de croissance et entrent en prolifration (Engelhardt, Kumar et al. 1997). A linverse, une dprivation en srum induit un arrt dans le cycle cellulaire et une diminution de lactivit tlomrase (Zhang, Piatyszek et al. 1996). Lors de cultures in vitro de CSH, on observe une augmentation importante de lactivit tlomrase pendant les deux premires semaines daction des cytokines, suivie dune diminution franche jusqu atteindre une valeur trs faible lorsque les cellules arrivent en diffrenciation terminale (Engelhardt, Kumar et al. 1997). La figure 9 rcapitule les donnes observes par diffrentes quipes sur le niveau dactivit tlomrase dans les cellules hmatopotiques. Pour ce qui est de la ligne mylode, on retrouve une activit tlomrase leve dans les progniteurs prcoces qui sont encore CD34+ (Hiyama, Hirai et al. 1995; Chiu, Dragowska et al. 1996; Sakabe, Yahata et al. 1998), puis une diminution au stade de progniteurs granulo/mono et rythro/mga (Chiu, Dragowska et al. 1996; Zhang, Piatyszek et al. 1996). Les monocytes et polynuclaires retrouvs dans le sang prsentent quant eux une activit tlomrase moyenne (Broccoli, Young et al. 1995). Les cellules hmatopotiques prsentant lactivit tlomrase la plus forte sont les prcurseurs lymphodes T prsents dans le thymus (Weng, Levine et al. 1996), qui vont en effet subir un grand nombre de divisions cellulaires. En revanche, les lymphocytes B et T prsents dans le sang, expriment peu de tlomrase, mais son niveau pourra tre augment lors de leur activation (Igarashi and Sakaguchi 1997).

39

Figure 9 : Niveau dactivit tlomrase dans les cellules hmatopotiques humaines. + - : activit trs faible, + : activit moyenne, ++ : activit forte, +++ : activit trs forte. Daprs (Ohyashiki, Sashida et al. 2002).

2.2. Statut des tlomres dans les cellules hmatopotiques Des observations montrant que lors de cultures liquides de progniteurs hmatopotiques en prsence de cytokines, on assiste un raccourcissement tlomrique (Vaziri, Dragowska et al. 1994), alors que lon sait que dans le mme temps lactivit tlomrase augmente, impliquent que la tlomrase ne suffit pas empcher la perte tlomrique pour ces cellules. De mme, lquipe de Lansdorp a montr que les tlomres des CSH de personnes adultes sont plus courts que ceux des CSH de sang de cordon ombilical, ce qui suggre une perte progressive de squences tlomriques au cours des divisions cellulaires (Vaziri, Dragowska et al. 1994). Il semblerait ainsi que lexpression de la tlomrase dans les cellules hmatopotiques permette de rduire et non dempcher lattrition tlomrique (Wang, Warner et al. 2005). La tlomrase pourrait donc jouer un rle dans linduction de la prolifration ou la protection des tlomres contre les

40

dommages lADN et non dans la prvention de lattrition tlomrique, ou bien seulement lorsque la taille des tlomres devient critique, pour viter lentre en snescence (Engelhardt, Wasch et al. 2004). 2.3. Pathologies constitutionnelles ou acquises lies la tlomrase touchant lhmatopose La dyskratose congnitale : La dyskratose congnitale est une maladie gntique rare caractrise par une dficience des tissus renouvellement rapide aboutissant un dfaut de pigmentation de la peau, une leucoplasie des muqueuses, une dystrophie des ongles puis une mort prmature due un appauvrissement de la moelle osseuse (diminution de lhmatopose) et lapparition de cancers. Il a t montr que les cellules de patients atteints par cette maladie prsentaient une forte instabilit gntique (pouvant favoriser lapparition de cancers) et un raccourcissement prcoce des tlomres (Mitchell, Wood et al. 1999). Les deux formes de cette maladie (forme rcessive lie au chromosome X ou forme autosomique dominante) sont lies des mutations dans les composants du complexe de la tlomrase : mutation dans le gne de la dyskrine pour la forme lie lX (Mitchell, Wood et al. 1999), et mutation des sous-units ribonuclique et catalytique de la tlomrase dans les formes autosomiques (Vulliamy, Marrone et al. 2001; Vulliamy, Walne et al. 2005). Les aplasies mdullaires : Dans 20% des aplasies mdullaires (insuffisance mdullaire due la disparition du tissu hmatopotique) ont t retrouves des mutations de la sous-unit ribonuclique hTR (Vulliamy, Marrone et al. 2002), ainsi que des mutations de la sous-unit catalytique hTERT dans un petit nombre dentre eux (Vulliamy, Walne et al. 2005). On peut observer de plus un raccourcissement important des tlomres chez les patients aplasiques (Lee, Kook et al. 2001).

3. Tlomrase et hmopathies malignes

Le statut dexpression de la tlomrase et la longueur des tlomres dans les hmopathies malignes ont t tudis dans de nombreux travaux. Les rsultats montrent

41

que, de manire gnrale, on retrouve des tlomres courts et une activit tlomrase augmente comme dans la plupart des tumeurs. Le tableau 2 rcapitule les donnes concernant lexpression de la tlomrase et la longueur des tlomres dans les hmopathies malignes (daprs (Ohyashiki, Sashida et al. 2002)).
Tableau 2 : Statuts des tlomres et de la tlomrase dans les hmopathies malignes

Hmopathie maligne Cellules hmatopotiques normales Leucmies aigus Mylode Lymphode Mylodysplasies Leucmie mylode chronique Phase chronique Phase blastique Leucmie lymphode chronique Phase prcoce Phase tardive Lymphome non Hodgkinien De type B De type T Maladie de Hodgkin
+/- +++ : niveau dexpression de la tlomrase.

Longueur des tlomres

Activit tlomrase +- ++

courts courts variable normaux courts courts normaux courts courts courts normaux courts

+++ +++ + ++ + + +++ + + ++ ++ + +

Daprs (Ohyashiki, Sashida et al. 2002), (Ohyashiki, Ohyashiki et al. 1997), (Ohyashiki, Ohyashiki et al. 1997) ; (Engelhardt, Mackenzie et al. 2000), (Engelhardt, Wasch et al. 2004), (Bock, Serinsoz et al. 2004), (Brousset, Chaouche et al. 1998).

3.1. Leucmies aigus La leucmie aigu (LA) est une hmopathie maligne caractrise par lexpansion dun clone leucmique bloqu un stade prcoce de diffrenciation, aboutissant une accumulation continue de cellules immatures dans la moelle puis dans le sang et donc

42

une forte insuffisance mdullaire. Cette hmopathie peut toucher aussi bien la ligne mylode que lymphode. La tlomrase est retrouve surexprime dans 75% des cas de leucmies aigus avec une surexpression jusqu dix fois par rapport aux cellules hmatopotiques normales. Certaines quipes placent lactivit tlomrase comme facteur pronostique puisque pour certains patients, le pronostic apparat moins bon lorsque lactivit tlomrase est leve (Ohyashiki, Ohyashiki et al. 1997). On peut dailleurs noter une diminution de lactivit tlomrase lorsque le patient est en rmission, mais elle peut augmenter nouveau des niveaux plus levs en cas de rechute (Tatematsu, Nakayama et al. 1996; Engelhardt, Mackenzie et al. 2000). Il est intressant de noter que les cellules tumorales comme par exemple dans les leucmies, prsentent des tlomres trs courts (Shay, Werbin et al. 1996; Franco, Ozkaynak et al. 2003), malgr une activit tlomrase leve. De plus, lanalyse de la longueur des tlomres de cellules de moelle osseuse ou de sang priphrique de patients leucmiques, permet trs souvent de dceler deux populations distinctes de cellules : lune avec des tlomres courts qui correspond au clone leucmique, lautre avec des tlomres plus longs qui correspond la population de cellules normales (Engelhardt, Ozkaynak et al. 1998; Franco, Ozkaynak et al. 2003). Ces donnes tendraient montrer que le clone leucmique puisse tre issu de cellules snescentes ayant subies des mutations dues linstabilit gntique, permettant une rexpression de la tlomrase. 3.2. Syndromes mylodysplasiques Les syndromes mylodysplasiques (SMD) sont des hmopathies clonales acquises caractrises par une hmatopose inefficace responsable de cytopnies priphriques contrastant avec une moelle osseuse de richesse normale. Cette maladie et le plus souvent accompagne dune forte instabilit gntique au niveau des cellules hmatopotiques avec une transformation frquente en leucmie aigu myloblastique. La mylodysplasie peut ainsi tre compare un tat pr-leucmique . Contrairement aux LA, la tlomrase est retrouve surexprime dans seulement 40% des SMD avec une augmentation de lordre de 2 5 fois (Ohyashiki, Ohyashiki et al. 1997; Engelhardt, Mackenzie et al. 2000; Bock, Serinsoz et al. 2004). On peut observer cependant un raccourcissement des tlomres dans plus de 60% des SMD, ce qui peut tre corrl avec linstabilit gntique et la prsence dun caryotype complexe correspondant

43

un facteur de mauvais pronostic (Sieglova, Zilovcova et al. 2004). Il a t montr une aggravation progressive de lattrition tlomrique entre les formes de SMD de bas grade et les leucmies (Mufti 2004), alors que lactivit tlomrase naugmente pas de faon drastique lors de laccutisation, puisque seulement 40% des LA secondaires ont une activit tlomrase leve. Cependant, au vu de lhtrognit des populations cellulaires mylodysplasiques, des tudes plus cibles sur les diffrentes populations permettront dapporter des prcisions quant au statut de la tlomrase et des tlomres dans les SMD et leur lien avec un facteur pronostique.

4. Rle de la tlomrase dans lhmatopose ?

Chez lhomme, le niveau dexpression de la tlomrase dans les cellules souches hmatopotiques semble tre extrmement rgul, et constitue un lment dterminant de lexpansion hmatopotique si on prend en compte le fait quune mutation de lun des lments du complexe tlomrasique peut conduire des aplasies mdullaires (Yamaguchi 2007). En effet, labsence de tlomrase fonctionnelle dans la cellule souche hmatopotique pourrait empcher ou ralentir lhmatopose normale. Rcemment, plusieurs tudes se sont intresses leffet de la modulation de la tlomrase dans les cellules hmatopotiques humaines en surexprimant hTERT dans des cellules souches CD34+ , ou en linhibant grce un dominant ngatif. Les rsultats obtenus montrent que le rle principal de la tlomrase dans les cellules hmatopotiques nest pas rellement le maintien de la longueur des tlomres puisque la surexpression de la tlomrase ne permet quune faible inhibition de lattrition tlomrique (Zimmermann, Glaser et al. 2004). En revanche, il apparat que la prsence dune activit tlomrase leve permette daugmenter la capacit dauto-renouvellement des cellules souches et de protger contre la mort par apoptose (Armstrong, Saretzki et al. 2005). Enfin, les diffrentes tudes tendent montrer que la tlomrase pourrait avoir une implication dans la diffrenciation hmatopotique. En effet, la surexpression de mTERT dans des cellules souches embryonnaires de souris, induit, en modifiant lexpression de nombreux gnes, une augmentation de leur diffrenciation vers les lignes hmatopotiques (Armstrong, Saretzki et al. 2005). Dautre part, linhibition de la tlomrase dans les cellules souches hmatopotiques humaines CD34+ a un effet dltre sur lapparition de colonies en milieu semi-solide notamment les CFU-GM (colony forming unit - granulo - monocytic),

44

et la surexpression dhTERT permet au contraire de favoriser lengagement vers la ligne rythrode, avec une augmentation de lapparition de BFU-E (burst forming unit erythroid) (Zimmermann, Glaser et al. 2004). Cette implication de la tlomrase dans la diffrenciation rythrode trouve un deuxime argument dans le fait que la prsence dune activit tlomrase leve et de tlomres longs dans les cellules souches hmatopotiques corrlent avec le potentiel prolifratif des progniteurs rythrodes long terme (Schuller, Jankowski et al. 2007).

45

46

PARTIE 2 : RESULTATS

47

RESULTATS

I. Objectifs du travail
Etant donn le rle jou par la tlomrase dans de nombreuses fonctions cellulaires telles que le maintien de la longueur des tlomres, la prolifration, la protection contre lapoptose ou la diffrenciation, de nombreuses tudes se sont intresses aux mcanismes molculaires de rgulation de cette enzyme. Il apparat ainsi que lactivit de la tlomrase dans les cellules est hautement rgule tous les niveaux (transcriptionnel et posttranscriptionnel). Cependant, on trouve peu de donnes bibliographiques dcrivant les stimuli intra ou extracellulaires pouvant interfrer avec les mcanismes de rgulation de lexpression dhTERT. Quelques rcentes tudes ont montr que la tlomrase pouvait tre rgule par des oncognes tel Her2/Neu (Goueli and Janknecht 2004), des facteurs de croissance comme lEGF (Epidermal Growth Factor) dans les cellules pithliales (Maida, Kyo et al. 2002), des cytokines avec par exemple lIL6 et lIGF-1 dans les cellules de mylome multiple (Akiyama, Hideshima et al. 2002), ou des hormones et notamment les strodes dans les cellules ovariennes (Kimura, Ohmichi et al. 2004), mais trs peu dtudes se sont penches sur le rle et la rgulation de la tlomrase au cours de lhmatopose. On sait que, dans les cellules souches hmatopotiques, lactivit de la tlomrase est trs faible et quelle peut augmenter en prsence dun mlange de cytokines permettant leur prolifration et leur diffrenciation (Engelhardt, Kumar et al. 1997). Par la suite, on observe une diminution graduelle de son activit lors de la fin de la maturation hmatopotique. Ces rsultats suggrent un rle des facteurs de croissance hmatopotiques dans la rgulation de la tlomrase, sans que les mcanismes de cette rgulation ne soient connus. Lobjectif gnral de notre tude a donc t dexplorer les voies de signalisation permettant la rgulation de la tlomrase par les facteurs de croissance hmatopotiques et certaines cytokines inhibitrices, aussi bien dans lhmatopose normale que leucmique.

48

Lobjectif de la premire partie de ltude, a t didentifier les voies de rgulation permettant linhibition de la tlomrase par le TNF- au niveau des cellules mylodes normales et leucmiques. En effet, cette cytokine inflammatoire inhibitrice de lhmatopose induit un phnotype de snescence dans les cellules hmatopotiques, avec une rosion des tlomres (Beyne-Rauzy, Recher et al. 2004). En deuxime partie nous avons ax notre tude sur les mcanismes et le rle de la rgulation de la tlomrase par deux cytokines majeures de lrythropose, lrythropotine (EPO) comme facteur activateur, et le TGF- comme cytokine inhibitrice.

49

II. Rsultats 1 : Rgulation de lexpression de la tlomrase par le TNFdans les cellules mylodes humaines normales et leucmiques
Article n 1 : Tumor necrosis factor alpha inhibits hTERT gene expression in human myeloid normal and leukemic cells.
Odile Beyne-Rauzy, Nas Prade-Houdellier, Ccile Demur, Christian Recher, Jacques Ayel, Guy Laurent, Vronique Mansat-De Mas. Blood 2005, Vol. 106 (9), p. 3200-3205.

Introduction
Le TNF- (Tumor Necrosis Factor alpha) est une cytokine existant sous forme transmembranaire (26kDa) et sous forme soluble (17kDa), scrte par plusieurs types cellulaires, dont les cellules du microenvironnement mdullaire. Il existe deux rcepteurs pour le TNF- , le TNFR1 (TNF Recepteur type 1, CD120a, p55/60) et le TNFR2 (TNF Recepteur type 2, CD120b, p75/80) faisant partie des rcepteurs de la superfamille du TNF qui sont caractriss par une organisation en homotrimres. Alors que le TNFR1 est ubiquitaire et peut tre activ par le TNF- membranaire ou soluble, le TNFR2 est retrouv uniquement sur les cellules immunitaires et endothliales et nest activ que par la forme membranaire du TNF- . De plus, le TNFR1 comporte sur sa partie cytoplasmique un domaine associ la transduction dun signal de mort cellulaire appel domaine de mort (death domain ou DD). Le systme TNF- et son rcepteur sont impliqus dans un large spectre deffets biologiques en fonction du type cellulaire, tels que des signaux de mort cellulaire, de prolifration, de diffrenciation, dinflammation et de tumorignse (Beutler 1990). Aprs la liaison du TNF- au TNFR1, les adaptateurs TRADD et TRAF2 sont rapidement recruts au recepteur via le DD, pour former un premier complexe (complexe

50

I), qui en induisant NF- B et les voies des MAP-Kinases va permettre lactivation de voies de survie cellulaire. Aprs internalisation de ce complexe et dissociation du TNFR1, le DD de TRADD peut intragir avec FADD et recruter la procaspase8 pour former le complexe II qui a quant lui un effet pro-apoptotique (Micheau and Tschopp 2003). Par ailleurs, le recrutement de la protine FAN au niveau du domaine NSD (N-SMase activating domain) du TNFR1, permet lactivation dune sphingomylinase et donc la gnration de cramide (Adam-Klages, Adam et al. 1996), qui va avoir un effet pro-apoptotique en agissant au niveau de la mitochondrie ou en activant la voie JNK (Morales, Lee et al. 2007) (Figure 10). TNF-

TNF-R1

TRADD TRAF2

ERK

p38

JNK NF-kB FLIP

TRADD FADD TRAF2 Caspase 8

Cer

JNK SURVIE Caspase 3

mitochondrie RLO

APOPTOSE Figure 10 : Principales voies de signalisation actives par le TNF- conduisant la survie cellulaire travers la formation du complexe I pro-survie, et lapoptose grce la formation du complexe II et lactivation des caspases, et la gnration de cramide. Cer : cramide, RLO : radicaux libres oxygns.

Comme vu plus haut, le TNF- est scrt par les cellules du microenvironnement mdullaire et va avoir un effet inhibiteur sur lhmatopose physiologique. Au niveau des cellules souches hmatopotiques, le TNF- a un rle dltre par la diminution de leur capacit dauto-renouvellement en favorisant leur diffrenciation (Dybedal, Bryder et al. 2001), mme si des tudes contradictoires ont montr que chez la souris il peut avoir un rle plutt positif des stades extrmement prcoces (Rogers and Berman 1994; Rebel, 51

Hartnett et al. 1999). Le TNF- a un effet inhibiteur sur les progniteurs prcoces en modulant lexpression des rcepteurs aux facteurs de croissance (Jacobsen, Rothe et al. 1994; Rusten, Jacobsen et al. 1994) ce qui aboutit un arrt de la prolifration et un blocage dans le cycle cellulaire. A un niveau plus tardif, il inhibe la formation des colonies des lignes granuleuse et rythrode (CFU-GM, BFU-E, CFU-E) lors de cultures en mthylcellulose (Johnson, Waddelow et al. 1989; Backx, Broeders et al. 1991; Caux, Favre et al. 1991; Means and Krantz 1993). Enfin, le TNF- induit de lapoptose et un arrt de maturation des prcurseurs rythrodes (De Maria, Zeuner et al. 1999). Des travaux prcdemment raliss au sein de lquipe ont montr que le TNFpouvait induire un tat de snescence dans la ligne leucmique KG1 lors dune exposition chronique long terme (7 14 jours) (Beyne-Rauzy, Recher et al. 2004). Cet tat de snescence est caractris par : un arrt de la prolifration avec blocage en phase G1 du cycle cellulaire, lapparition de Senescence-associated- -galactosidase (SA- -gal), le raccourcissement des tlomres, ceci accompagn dune importante instabilit chromosomique. Linduction de la snescence par le TNF- a pu tre confirm sur des cellules de moelle de patients atteints de leucmie aigue mylode (LAM), ainsi que sur des cellules hmatopotiques normales. Enfin, lajout de GM-CSF dans le milieu (cytokine stimulant la ligne mylode), permet de protger contre leffet du TNF- et mme dchapper ltat de snescence lorsquil est ajout aprs 14 jours de culture en prsence de TNF- . Cet tat de snescence est prcd par une diminution de lexpression du gne hTERT partir de 2 heures induite par le TNF- , et une diminution de lactivit de la tlomrase partir de 7 jours. Lobjectif de ltude qui suit tait donc de dterminer les mcanismes de rgulation transcriptionnelle de hTERT par le TNF- , permettant daboutir un phnotype de snescence lors dune exposition chronique.

Rsultats
La ligne KG1 sur laquelle nous avons mene cette tude est une ligne leucmique de phnotype prcoce (prsence du marqueur CD34+), qui prsente une rsistance de nombreux traitements. Dans un premier temps, nous avons montr que le TNF- inhibait hTERT non seulement dans la ligne KG1, mais galement dans des cellules hmatopotiques normales et leucmiques. Afin de reproduire une hmatopose in vitro, nous avons utilis

52

des cellules souches CD34+ obtenues par tri positif sur colonne aimante, partir des cellules de la moelle osseuse issues de fragments dos spongieux de sujets sains. Les cellules ont t mises en culture en prsence dun mlange de cytokines (stem cell factor [SCF], FLT3-L, Interleukine-3 [IL3] et Thrombopotine [TPO]) permettant dobtenir une diffrenciation mylode granuleuse. La diffrenciation est suivie grce la disparition du marqueur CD34, lapparition de marqueurs de diffrenciation CD38 et CD33, ainsi que par le changement de morphologie des cellules (Figure 11). On peut remarquer que lactivit tlomrase au dpart trs faible (dans les cellules CD34+), augmente fortement pour atteindre un pic 7 jours de culture, puis diminue lentement lors de la suite de la diffrenciation (Figure 11 et Fig. 1D p. 3202) comme dj observ par Engelhardt et al. (Engelhardt, Kumar et al. 1997). Lajout de TNFdans cette culture induit un ralentissement de la prolifration (Fig. 1A p.3202), une acclration nette de la diffrenciation avec une perte plus prcoce du marqueur CD34 (Fig. 1B p.3202), et une inhibition de la transcription du gne hTERT, qui se traduit par une inhibition au moins partielle (50% dinhibition) de lactivit de la tlomrase partir de 7 jours de culture (Fig. 1C et 1D p.3202).
IL3, SCF FLT3-L, TPO IL3, SCF FLT3-L, TPO

J0

J7

J14

Ig-PE CD34-PE

CD34

CD34-PE

CD34-PE

CD34-PE

Activit tlomrase

J0

J7

J14

Figure 11 : Diffrenciation in vitro des cellules souches hmatopotiques CD34+, volution de lactivit tlomrase.

53

En parallle, des cellules issues de moelle de patients atteints de LAM ont t mises en culture et traites par du TNF- . Ce traitement a pour effet dinhiber la transcription du gne hTERT ds 2h ou 6h dexposition, dans la majorit des chantillons (Tableau 1 p. 3202). Ces deux rsultats obtenus sur des cellules hmatopotiques normales et leucmiques confirment que le TNF- inhibe hTERT au niveau transcriptionnel. La suite du travail a t consacre la caractrisation des voies de signalisation actives par le TNF- permettant linhibition de la transcription du gne hTERT dans le modle KG1. Parmi les voies susceptibles dtre impliques dans notre modle, nous avons envisag le rle du cramide, second messager lipidique impliqu dans la signalisation du TNF- . En effet, dans la ligne leucmique U937, le TNFinduit la gnration de cramide endogne en activant une sphingomylinase (SMase) responsable de lhydrolyse de la sphingomyline membranaire (Obeid, Linardic et al. 1993) avec pour effet linduction dapoptose. Dautre part, dans une ligne dadnocarcinome pulmonaire humain, le cramide permant, la surexpression de la SMase bactrienne ou lexposition la Daunorubicine gnrant du cramide endogne, peuvent induire une inhibition de hTERT et un arrt de la prolifration des cellules (Ogretmen, Schady et al. 2001). Sur la base de ces observations, nous avons mis lhypothse que, dans la ligne KG1, le TNFrgule ngativement hTERT en gnrant du cramide. La production de cramide dans la cellule a t quantifie grce un marquage avec de lacide palmitique triti, extraction des lipides, sparation par chromatographie couche mince et mesure de la radioactivit. On peut ainsi observer une augmentation de la production de cramide sous laction du TNFavec un pic 3 minutes de traitement (Fig. 3A p.3203). De plus, il apparat quun traitement des cellules au cramide permant induit une inhibition de la transcription du gne hTERT (Fig. 3B p.3203), suggrant que ce second messager lipidique soit responsable de cette inhibition. Des tudes antrieures ont montr que le TNF- pouvait induire de nombreuses voies de signalisation dont les voies des MAP-Kinases qui jouent un rle important dans la rgulation de nombreux gnes (van Horssen, Ten Hagen et al. 2006). De plus, le TNFpeut induire la voie JNK grce la gnration de cramide par une SMase (Verheij, Bose et al. 1996). Nous avons ainsi recherch une implication de ces voies en aval du cramide, et nous avons pu montrer que dans le modle KG1, les voies p38MAPK et JNK (mais pas ERK) taient actives de faon prcoce par le TNF- (Fig. 2 p.3203). Cependant, seule la

54

voie JNK semble tre implique dans la rgulation de lexpression du gne hTERT, comme montr dans le Tableau 2 p.3203, grce lutilisation dun inhibiteur de cette voie. De plus, linhibition de la voie JNK bloque la rgulation ngative du gne hTERT induite par le cramide permant (Fig. 3B p.3203). Enfin, le GM-CSF, qui avait un effet protecteur en ralentissant lentre en snescence induite par le TNF- , agit la fois de faon trs prcoce en bloquant le pic de production de cramide, ainsi quen aval du cramide, pour contrer linhibition transcriptionnelle de la tlomrase (Fig. 4 p.3204).

Conclusion et perspectives
En conclusion, cet article a permis de montrer quau cours de lhmatopose normale ou leucmique, la tlomrase pouvait tre rgule ngativement par une cytokine inhibitrice, le TNF- , et positivement par un facteur de croissance, le GM-CSF (Figure 12).

Cramide

Figure 12 : Schma daction du TNFet du GM-CSF sur la transcription du gne hTERT, dans la ligne KG1.

Certaines publications ont montr que le TNF- avait au contraire un effet positif sur lactivit tlomrase en permettant sa translocation au noyau grce linteraction avec NF- B (Akiyama, Hideshima et al. 2003) (Akiyama, Yamada et al. 2004). Cependant, ces

55

travaux ne sont pas en dsaccord avec nos rsultats puisque cette activation de la tlomrase intervient de faon trs prcoce (60 minutes), soit avant leffet ngatif sur la transcription du gne (2 heures) que nous avons pu observ dans notre modle. Nos travaux ont permis de mettre en vidence limplication de la voie cramideJNK dans la rgulation ngative de la tlomrase par le TNF- dans la ligne KG1. Le cramide est largement connu comme mdiateur lipidique impliqu dans la cascade de signalisation du TNF- . Cependant, des travaux antrieurs raliss dans notre quipe avaient montr que, contrairement la ligne U937 o le TNF- induit de lapoptose, dans la ligne KG1a, il est responsable dune augmentation de la prolifration et ninduit pas de gnration de cramide (Bettaieb, Record et al. 1996). Cependant, le cramide permant rtablit lapoptose dans cette ligne, suggrant que cest la gnration de cramide qui est bloque et non son effet. Ceci peut tre expliqu par une indisponibilit du substrat (la sphingomyline hydrolysable) (Bettaieb, Record et al. 1996), une augmentation de lactivit des PKC (Mansat, Laurent et al. 1997), ou une augmentation des dfenses antioxydatives (Mansat-de Mas, Bezombes et al. 1999) ayant un effet inhibiteur sur la SMase. En ce qui concerne la tlomrase, nous avons pu constater que le TNF- na pas deffet inhibiteur sur le transcrit hTERT, ni dans les cellules KG1a (ce qui corrle avec la TNFU937
SURVIE PROLIFERATION

Cer

hTERT

SENESCENCE

APOPTOSE

TNFKG1a
SURVIE PROLIFERATION
APOPTOSE

Cer
hTERT
SENESCENCE

TNFKG1
SURVIE PROLIFERATION APOPTOSE

Cer
hTERT

SENESCENCE

Figure 13 : Schma hypothtique expliquant les diffrents effets du TNFleucmiques.

dans trois lignes

56

non-induction de cramide), ni dans les cellules U937 (rsultats non montrs). Ainsi, le TNF- induit une rponse diffrente en fonction de la ligne cellulaire : (i) lapoptose dans la ligne U937 avec prdominance de la voie du cramide, (ii) la prolifration dans la ligne KG1a, la gnration de cramide tant bloque, (iii) la snescence dans la ligne KG1 (comme montr prcedemment (Beyne-Rauzy, Recher et al. 2004)), avec inhibition de hTERT (Figure 13). Dans ce dernier cas, on nobserve pas dapoptose induite par le TNF- malgr la gnration de cramide, ceci pouvant tre expliqu par la surexpression de la protine anti-apoptotique Bcl-2 retrouve dans cette ligne (Quillet-Mary, Mansat et al. 1996), qui bloque leffet pro-apoptotique du cramide (Allouche, Bettaieb et al. 1997). De faon intressante, une tude a montr que la surexpression de Bcl-2, en bloquant la signalisation apoptotique, oriente les cellules vers la snescence (Schmitt, Fridman et al. 2002), comme retrouv dans la ligne KG1. La voie JNK a dj t retrouve en aval du cramide dans la signalisation du TNF(Verheij, Bose et al. 1996), mais son implication dans la rgulation ngative de la tlomrase va lencontre de certaines publications. En effet, dans les cellules pithliales, la voie JNK permet au contraire lactivation de la transcription du gne hTERT par le facteur AP-1 (Alfonso-De Matte, Yang et al. 2002), mme si AP-1 a plus rcemment t dcrit comme facteur inhibiteur de la transcription du gne hTERT (Takakura, Kyo et al. 2005). Des tudes plus approfondies permettront de dtailler cette voie, notamment pour connatre le facteur de transcription impliqu. Il est nanmoins possible de proposer des facteurs de transcription potentiels, tel que AP-1, c-myc, qui peut tre considr comme le plus important rgulateur transcriptionnel du gne hTERT, et qui peut tre cliv en aval de la voie du cramide (Ogretmen, Kraveka et al. 2001) ou Sp-1 dont laction est de mme bloque par le cramide (Wooten and Ogretmen 2005). Nos observations confrent un rle protecteur au GM-CSF vis--vis du TNF- , en agissant la fois en amont de la gnration de cramide et en aval au niveau de la cascade de signalisation, pour bloquer linhibition de la tlomrase, sans que les mcanismes de son action ne soient connus. Ce rsultat est en accord avec les travaux de Kelly et coll. dans lesquels le GM-CSF inhibe la gnration de cramide induite par les rayonnements ionisants, par lintermdiaire de la PKC (Kelly, Tang et al. 1998). Cependant, aucune

57

PKC nest implique dans notre modle tant donn que les inhibiteurs de PKC (calphostine et chlrythrine) ne bloquent pas leffet du GM-CSF (rsultats non montrs). Les rsultats obtenus ont t retrouvs dans des cellules de patients atteints de LAM, o le TNFinduit une diminution variable du transcrit hTERT dans les 10 induit une inhibition de la prolifration, une chantillons de moelle de patients tests. En ce qui concerne les progniteurs hmatopotiques normaux, le TNFacclration de la diffrenciation ainsi quun blocage de la transcription du gne hTERT. Cette inhibition de la transcription du gne induit une diminution plus rapide de lactivit tlomrase dans les cellules traites par le TNF- , partir de 7 jours de culture. La tlomrase peut tre implique dans diffrentes fonctions cellulaires, dont la protection contre lattrition tlomrique, la prolifration court et long terme, la protection contre lapoptose, ou la diffrenciation. Dans notre modle de diffrenciation mylode, son rle na pas encore t dtermin mme sil semble quelle permette de maintenir lexpansion cellulaire en vitant lentre en snescence. En effet, ltude prcdemment mene par le laboratoire montrait que linhibition de la tlomrase par le TNF- conduisait un phnotype de snescence dans les cellules KG1, avec inhibition de la prolifration, apparition de SA- -galactosidase, et attrition tlomrique (Beyne-Rauzy, Recher et al. 2004). Il faut noter de plus que ce phnotype de snescence est accompagn dune importante instabilit chromosomique avec apparition de multiples anomalies. Ces rsultats peuvent tre corrls certains travaux montrant quau cours de maladies inflammatoires du tube digestif (o le TNFest en excs), on observe un raccourcissement des tlomres et une instabilit chromosomique (O' Sullivan, Bronner et al. 2002) ce qui permet de faire le lien entre contexte inflammatoire et risque de carcinogense. Dans notre modle hmatologique, on peut donc envisager que la prsence de TNFen excs dans lenvironnement mdullaire au cours des syndromes mylodysplasiques, puisse en inhibant la tlomrase, induire un phnotype snescent accompagn dinstabilit gntique, pouvant favoriser lmergence du clone leucmique.

58

III. Rsultats 2 : Rle et rgulation de la tlomrase dans les progniteurs rythrodes humains

Article n2 : Human telomerase is regulated by erythropoietin and transforming growth factor- in human erythroid progenitor cells
Nas Prade-Houdellier, Elise Frbet, Ccile Demur, Emilie-Fleur Gautier, Franois Delhommeau, Anne-Lise Bennaceur-Griscelli, Clment Gaudin, Vincent Martinel, Guy Laurent, Vronique Mansat-De Mas, Odile Beyne-Rauzy Leukemia (2007), 21 : 2304-2310

Introduction
Le travail prcdent a permis de montrer que lactivit tlomrase, trs faible dans les cellules souches hmatopotiques CD34+, augmentait en prsence de cytokines au dbut de la culture pour diminuer nouveau en fin de diffrenciation (Beyne-Rauzy, Prade-Houdellier et al. 2005), comme dj publi par Engelhardt et coll. (Engelhardt, Kumar et al. 1997). De plus, une cytokine inflammatoire inhibitrice de lhmatopose, le TNF- , inhibait la tlomrase tout en acclrant la diffrenciation (Beyne-Rauzy, PradeHoudellier et al. 2005). Lensemble de ces rsultats nous ont permis dmettre lhypothse selon laquelle la tlomrase joue un rle critique dans lexpansion hmatopotique, et peut donc tre rgule par diffrents facteurs de croissance ou cytokines rgulatrices de lhmatopose. Trs peu de travaux ont explor les voies de rgulation de la tlomrase par les cytokines de lhmatopose. Nous nous sommes plus particulirement intresss la

59

rgulation de la tlomrase dans lrythropose. Ce modle nous a en effet paru intressant tout dabord car cette ligne rythrode est rgule principalement par un facteur de croissance, lrythropotine (EPO), qui joue un rle majeur dans lexpansion rythrode. Dautre part, plusieurs cytokines inflammatoires telles que le TGF- (mais aussi le TNF- , linterfronou le systme Fas/FasL), ont un effet dltre sur lrythropose, et peuvent participer lapparition danmies acquises chez les patients atteints de syndromes mylodysplasiques (SMD), dinflammations chroniques ou de cancers. LEPO est une hormone glycoprotique de 30 kDa produite par les reins en rponse un stress hypoxique. Cest le facteur de croissance principal de la ligne rouge, il induit la prolifration des progniteurs rythrodes, permet de les protger contre lapoptose, et dinduire leur diffrenciation afin daugmenter la quantit dhmaties dans le sang. En se fixant son rcepteur (superfamille des rcepteurs aux immunoglobulines), lEPO permet un changement conformationnel de celui-ci qui induit le rapprochement de deux tyrosines kinases JAK2 (Janus Kinase 2) qui se transphosphorylent et phosphorylent le rcepteur permettant lactivation de nombreuses voies de signalisation. Parmi ces voies, les principales sont la voie JAK2/STATs, la voie PI3Kinase/Akt, et la voie des MAPKinases (Richmond, Chohan et al. 2005) (Figure 14) qui aboutissent la prolifration, la protection contre lapoptose et la diffrenciation.

P P

P P

MAPK

STATs

PI3K Akt
Figure 14 : des en Mcanismes voies aval de de dactivation signalisation lEPO.

60

Le TGF- , est une protine homodimrique scrte dans un contexte inflammatoire, qui joue un rle dans la prolifration, la diffrenciation, la migration cellulaire et lapoptose. Laction du TGFpasse par lactivation dune cascade de signalisation impliquant les protines Smads, qui forment une famille de facteurs de transcription agissant en complexe afin dactiver ou dinhiber la transcription de certains gnes (ten Dijke and Hill 2004). Leffet du TGF- sur lrythropose est plutt dltre car il acclre la diffrenciation mais inhibe la prolifration des progniteurs rythrodes (Zermati, Fichelson et al. 2000) avec pour effet un appauvrissement de la ligne rouge pouvant conduire lapparition danmies, sans que les mcanismes de cette inhibition ne soient connus. Lhypothse de ce travail est donc que la tlomrase peut tre rgule de faon antagoniste par lEPO et le TGF- lors de lrythropose, et que cette enzyme joue un rle dans lexpansion rythrode.

Rsultats
Effet de lEPO sur la rgulation de la tlomrase. Cette tude a t ralise principalement sur le modle mgacaryoblastique UT7, qui cultive en prsence dEPO prsente un phnotype de progniteur rythrode. Cette ligne tant dpendante de lEPO, afin dtudier la signalisation induite par cette cytokine, nous avons ralis des expriences de sevrage de 16 heures puis restimulation par lEPO (figure 15 A). Comme on peut le voir sur la figure 15, le sevrage en EPO induit un arrt de la prolifration ds 16 heures (dtecte par incorporation de BrdU, figure 15 C et D), et entrane la mort des cellules en 48 heures (figure 15 B). Lors de la stimulation par lEPO, on observe une reprise totale de la prolifration partir de 48 heures de culture comme en tmoigne laugmentation de lincorporation de BrdU (figure 15C et D).

61

B
100 90 80 70 60 50 40 30 20 10 0 0 1 2 3 4 5 6 7

C
$%$& " #
0 1 2 10 5 0

20 18 16 14 12 10 8 6 4 2 0 -16h 0 6h 24h 48h

D -16h
BrdUFITC

24h

48h

Iodure de propidium Figure 15 : Effet de lEPO sur la prolifration des cellules UT7. A) Principe des expriences de sevrage/stimulation par lEPO. B) Courbe de viabilit des cellules UT7 aprs sevrage en EPO et restimulation (courbe bleu) ou non (courbe noir) par lEPO. C) Pourcentage de cellules ayant incorpor du BrdU avant sevrage (-16h), aprs sevrage (0) et aprs stimulation par lEPO (6 48h). D) Marquage BrdU-FITC/Iodure de propidium des cellules UT7 lors dun sevrage et restimulation par lEPO.

La question pose tant la rgulation de la tlomrase par lEPO, nous avons test cette hypothse dans le modle de sevrage et restimulation par lEPO des cellules UT7. Ainsi que le montre la figure 1 de larticle, le sevrage induit une diminution de la quantit de transcrit hTERT dans la cellule, mesure par RT-PCR quantitative, mais pas de lactivit tlomrase car la demi-vie de la protine est longue (probablement suprieure 24 heures). De la mme faon, lajout dEPO dans le milieu induit une augmentation de la

62

transcription du gne hTERT ds 6 heures, suivie par une augmentation de lactivit tlomrase 24 heures (figure 1a et 1b de larticle). Ces rsultats montrent que lEPO rgule transcriptionnellement la tlomrase. Rle de la tlomrase dans la signalisation induite par lEPO. Etant donn leffet connu de lEPO sur la prolifration et la survie des cellules rythrodes, nous nous sommes intresss au rle de linduction de la tlomrase dans ce modle. Pour cela, lactivit de la tlomrase dans les cellules UT7 a t inhibe par lexpression dun dominant ngatif de la tlomrase, obtenue par infection rtrovirale des plasmides Mig-R contrle et Mig-R-DN-hTERT (obtenus en collaboration avec le Dr. F. Delhommeau, INSERM U362, Villejuif) (figure 16). Malgr linhibition effective de la tlomrase, on ne note pas de ralentissement prcoce de la prolifration en rponse lEPO (pendant les 15 premiers jours de culture), ni de sensibilisation lapoptose dans les cellules UT7-DN-hTERT (figure 2 de larticle). Cependant, une tude dj publie avait montr qu partir de 4 semaines, lexpression du DN-hTERT dans les cellules UT7 avait pour effet dinduire un raccourcissement tlomrique et un arrt de la croissance cellulaire (Delhommeau, Thierry et al. 2002), indiquant un rle de la tlomrase dans la prolifration long terme.

5LTR Ori Mig-R Amp 3LTR

Ori IRES Amp

5LTR Mig-RDN-hTERT DNhTERT

GFP 3LTR GFP IRES

Figure 16 : Schma des constructions plasmidiques pour linfection rtrovirale.

Etude de la signalisation permettant la rgulation de la tlomrase par lEPO. Pour la suite de ltude, nous avons cherch dterminer les voies de signalisation impliques dans la rgulation de la tlomrase par lEPO dans le modle UT7. Comme nous lavons vu, le gne de la tlomrase peut tre rgul par de nombreux facteurs transcriptionnels dont les principaux sont c-myc et sp1. De plus, lEPO peut stimuler la transcription du gne c-myc (Chen and Sytkowski 2001). Leur implication dans la voie de 63

rgulation de la tlomrase par lEPO a donc t tudie. Nous avons tout dabord montr par gel retard que lEPO induisait la fixation de c-myc sur le promoteur dhTERT, grce une sonde comportant la squence de fixation de c-myc (E-box) complte par son environnement direct sur le promoteur du gne hTERT (figure 3a de larticle). En revanche, la fixation du facteur de transcription sp1 sur le promoteur du gne hTERT nest pas module par lEPO (rsultats non montrs). De plus, lEPO semble induire la transcription du gne c-myc et donc lexpression de la protine (figure 3b et c de larticle) dans le modle UT7. Afin de confirmer la contribution de c-myc dans la rgulation transcriptionnelle dhTERT par lEPO, nous avons inhib cette protine dans la ligne UT7 par une stratgie de siRNA. Les rsultats obtenus montrent quen labsence de protine cmyc dans la cellule, le traitement par lEPO ne parvient pas induire la transcription du gne hTERT. Lensemble de ces rsultats montre que la protine c-myc est le facteur de transcription reliant lEPO la tlomrase. En se liant son rcepteur, lEPO active de nombreuses voies de signalisation, dont les trois principales sont la voie JAK2/STAT5, la voie des MAPKinases, et la voie PI3Kinase/Akt, qui aboutissent linduction de la prolifration et de la survie cellulaire. Limplication de ces voies dans la rgulation de c-myc et de la tlomrase par lEPO a t teste grce lutilisation dinhibiteurs pharmacologiques. Nous avons ainsi pu montrer que la voie JAK2/STAT5 participait linduction de lexpression de c-myc, sa fixation sur le promoteur du gne de la tlomrase, et donc la rgulation transcriptionnelle de la protine hTERT (Figure 4 de larticle). Effet dun inhibiteur de lrythropose, le TGF- , sur la rgulation de hTERT. Le TGF- a, contrairement lEPO, un effet plutt dltre sur lrythropose (Zermati, Fichelson et al. 2000) et bloque la prolifration des cellules UT7 (figure 17).
20 18 16 14 12 10 8 6 4 2 0 0
+ Epo +Epo +TGF-beta

Nombre de cellules .10

Figure 17 : Courbe de viabilit reprsentant leffet du TGFsur lexpansion de la ligne UT7.

2 4 6 8 10

64

Dautre part, quelques tudes ont montr que le TGF - pouvait induire une inhibition de la transcription du gne de la tlomrase, en agissant au niveau de c-myc (Rama, Suresh et al. 2003; Li, Xu et al. 2006). Nous avons donc test son influence sur la rgulation de hTERT par lEPO dans notre modle rythrode UT7. Lajout de TGF- dans le milieu de culture lors de la restimulation des cellules UT7 par lEPO, bloque linduction transcriptionnelle de la tlomrase (figure 5a de larticle). Cependant, le traitement par le TGF- , na aucune influence sur lactivation de la voie JAK2/STAT5 (rsultats non montrs), ni sur linduction de lexpression de la protine c-myc ou sa fixation sur le promoteur du gne hTERT (figure 18 A et B). Daprs les travaux raliss par Hu et coll. chez le rat, ainsi que A
+Epo +Epo+TGF-

B
0
complexe de fixation de c-myc

+Epo

+Epo+TGF-

1h

2h

1h 2h

ite e i n p t t ro o m -p +c

ur

0 c-myc actine

2h

6h

2h

6h

Figure 18 : Effet du TGF- sur c-myc. A) Effet du TGF- (10ng/ml) sur linduction de la protine c-myc par lEPO dans la ligne UT7 (western blot sur extraits nuclaires). B) Effet du TGF- sur la fixation de cmyc sur le promoteur dhTERT induite par lEPO (gel retard).

Li et coll. chez lhomme, le TGF- peut agir directement sur le promoteur du gne hTERT travers lactivation de la voie des Smads (Hu, Tack et al. 2006; Li, Xu et al. 2006). On peut donc formuler lhypothse selon laquelle le TGFactiverait une cascade de signalisation aboutissant linduction de la protine Smad3, qui en se fixant sur le promoteur du gne hTERT sans empcher la fixation de c-myc sur la E-box, inhiberait la transcription du gne, et donc lexpression de la tlomrase. Lutilisation dun si-RNA ciblant Smad3 a permis la dpltion de cette protine, et on observe dans ce contexte un blocage de leffet inhibiteur du TGF- sur linduction transcriptionnelle du gne hTERT par lEPO dans les cellules UT7 (figure 5 de larticle). Smad3 semble donc tre implique dans la rgulation transcriptionnelle de la tlomrase, mais son mode daction reste encore lucider.

65

Rgulation de la tlomrase par lEPO et le TGFrythrodes.

dans les progniteurs

Lobjet de la dernire partie de ltude a t dtudier la rgulation de la tlomrase par lEPO et le TGF- dans des cultures primaires de progniteurs rythrodes humains. Pour cela, nous avons ralis des cultures de cellules souches hmatopotiques CD34+ multipotentes in vitro, isoles partir de fragments dos spongieux issus de chirurgie de la hanche (en collaboration avec le Service de chirurgie orthopdique du CHU de Purpan Toulouse). Les cellules mononucles extraites de la moelle osseuse sont tout dabord isoles grce un gradient de densit, puis marques par un anticorps anti-CD34 coupl

A
" # "

%' #( $

"

#)*

)*%$ , - .

2
1

%(

!$ )/*(%$ 0

$%&

B
CD34+ CD36+
CFU-E Erythroblaste basophile

SCF IL3 IL6

+ Epo
CFU-GEMM BFU-E Erythroblaste acidophile

Cellule souche

J0

, -.

J7 = tri CD36+

J15

CD34 CD36 GPA

Figure 19 : Protocole de diffrenciation rythrode in vitro. A) Obtention des cellules souches hmatopotiques CD34+ partir de fragments osseux de la hanche et tri sur colonne magntique. B) Diffrenciation rythrode in vitro : premire phase sans EPO, deuxime phase avec ajout dEPO aprs le tri des cellules CD36+.

66

des billes magntiques, en vu dune purification positive sur colonne aimante. Cette technique permet dobtenir des cellules CD34+ pures plus de 80% (figure 19 A). Les cellules sont ensuite mises en culture suivant le protocole tablit par Freyssinier et collaborateurs, permettant dobtenir une diffrenciation rythrode complte in vitro (Freyssinier, Lecoq-Lafon et al. 1999). Cette culture se droule en deux phases, la premire permet lobtention de progniteurs rythro-mgacaryocytaires grce laction de trois cytokines, le SCF, lIL3 et lIL6. Au bout de 7 jours de culture, les cellules sont tries par slection positive sur colonne aimante, sur la base de lexpression du marqueur rythrode prcoce CD36. Les cellules obtenues sont ensuite rcupres pour une deuxime phase de culture en prsence des trois cytokines plus lEPO, afin de permettre la diffrenciation rythrode terminale, que lon peut suivre grce lapparition de la glycophorine A (GPA) la surface des cellules (figure 19 B). Nous avons tout dabord voulu vrifier limpact de lEPO sur la transcription du gne hTERT dans les progniteurs rythrodes. Pour cela, les cellules CD36+ tries ont t sevres totalement (aucune cytokine dans le milieu) pendant 4h puis restimules par lEPO uniquement. Dans ce cadre, on observe une augmentation de la transcription du gne hTERT aprs ajout dEPO, avec un blocage de cette induction en prsence dinhibiteur de la voie JAK2/STAT (figure 6 de larticle). Ces rsultats confirment leffet de lEPO sur la rgulation de la tlomrase dans les progniteurs rythrodes humains, avec implication de la voie JAK2/STAT, comme dans la ligne UT7. Lajout de TGF- en mme temps que lEPO dans la deuxime phase de culture A
30 +Epo +Epo +TGF-beta
% de cellules GPA+
100 90 80 70 60 50 40 30 20 10 0 0 1 2 3 4 5 6 7 8

B
+Epo +Epo +TGF-beta

Nombre de cellules .10 6

25 20 15 10 5 0 0 1 2 3 4 5 6 7 8

Jours aprs ajout dEPO

Jours aprs ajout dEPO

Figure 20 : Effet du TGF- sur la prolifration (A) et lapparition du marqueur GPA (B) sur les cellules rythrodes lors de la deuxime phase de culture en prsence dEPO.

67

induit une forte inhibition de la prolifration des cellules rythrodes (figure 20 A), ainsi quune acclration nette de la diffrenciation comme le suggre laugmentation rapide du nombre de cellules GPA+ (figure 20 B). Ces rsultats sont en accord avec une tude dj publie par Zermati et coll. (Zermati, Fichelson et al. 2000), en revanche, aucune tude na t mene sur leffet du TGFsur la tlomrase dans ce systme de diffrenciation rythrode. Lors de lajout dEPO au dbut de la deuxime phase de culture, on observe une augmentation de lactivit tlomrase en rapport avec laugmentation du transcrit du gne hTERT, puis lactivit tlomrase diminue au cours de la diffrenciation terminale (figure 21). La prsence de TGF- bloque lactivation de la tlomrase par lEPO, aussi bien au niveau du transcrit (figure 6 de larticle), quau niveau de son activit, et induit donc une diminution plus rapide de lactivit tlomrase lors de la culture (figure 21).

1,8 1,6

Epo Epo + TGF

Activit tlomrase

1,4 1,2 1,0 0,8 0,6 0,4 0,2 0,0 0 2 4 6 8

Figure 21 : Effet du TGFsur lactivit tlomrase lors de la deuxime phase de diffrenciation rythrode. Jours aprs ajout dEPO

Conclusion et perspectives
Cette tude montre pour la premire fois que lEPO peut rguler

transcriptionnellement hTERT travers lactivation de la voie JAK2/STAT et du facteur de transcription c-myc. Dautre part, le TGF- , qui a un effet inhibiteur sur lrythopose, induit quand lui la protine Smad3, qui bloque lactivation transcriptionnelle de la tlomrase par lEPO (figure 22). Les rsultats obtenus sur la ligne UT7 ont, de plus, pu tre confirms sur des cultures primaires de progniteurs rythrodes humains.

68

REPO

+ 34
'

Figure 22 : Schma de

) -

rgulation de la transcription du gne hTERT par lEPO et le TGF- .

'

Limplication de la voie JAK2/STAT5 dans la rgulation de la tlomrase par lEPO trouve son importance dans le contexte de la rcente dcouverte de la mutation V617F de la protine JAK2 dans la polyglobulie primitive (James, Ugo et al. 2005). Cette mutation induit en effet une activation constitutive de JAK2 et donc des voies de signalisation en aval du rcepteur lEPO. Dans ces conditions, on peut supposer que la prsence de JAK2V617F dans la cellule conduise une activation de la tlomrase mme en labsence dEPO, ce qui contribuerait lexpansion rythrode. Les rsultats montrant la prsence dune activit tlomrase importante dans les granulocytes issus de moelles de polyglobulies primitives, appuient cette hypothse (Ferraris, Mangerini et al. 2005). La prsence de c-myc comme facteur de transcription impliqu dans la rgulation de la tlomrase par lEPO tait en partie attendue tant donn quil sagit de lun des facteurs de transcription les plus importants du gne hTERT, et quil peut tre retrouv en aval de lEPO (Chen and Sytkowski 2001). En revanche, dans cette dernire tude, la rgulation de c-myc par lEPO impliquait les voies MAPK et PI3K, ce qui nest pas le cas dans notre modle. Cependant, cette tude a t mene dans des cellules murines, et il a t montr par ailleurs dans un modle humain, quen rponse lIL2, STAT5 active transcriptionnellement lexpression de c-myc (Lord, McIntosh et al. 2000).

69

Nos observations montrent par ailleurs que le TGF- a un effet dominant sur lEPO en contrant son action sur la tlomrase. Ce rsultat trouve son importance dans certaines pathologies telles que les cancers, les maladies inflammatoire ou les mylodysplasies, dans lesquelles le taux de TGF- est anormalement lev, et o lon retrouve de nombreux cas danmies. Dans notre modle, le mode daction du TGFsur la rgulation de la induit tlomrase est indpendant de la voie induite par lEPO. En effet, le TGF-

lactivation de la protine Smad3, mais na pas deffet sur lactivation de la voie JAK2/STAT5, ni sur lexpression de la protine c-myc ou sa fixation sur le promoteur dhTERT. Ce rsultat est en dsaccord avec ltude publie par Rama et coll. montrant que dans le trophoblaste, le TGFinhibe la transcription du gne hTERT en inhibant lexpression de la protine c-myc (Rama, Suresh et al. 2003). Cependant, une tude plus rcente montre que dans le cancer du sein, le TGF- inhibe lexpression de la tlomrase de manire prcoce, sans inhiber lexpression de c-myc ni sa fixation sur le promoteur dhTERT (Li, Xu et al. 2006). Dans cette dernire tude, les auteurs proposent un modle selon lequel la protine Smad3 active par le TGF- se fixe sur la protine c-myc sur le promoteur dhTERT, ce qui a pour effet de bloquer la transcription du gne. Cette hypothse pourrait tre envisage dans notre modle, mme si des expriences prliminaires ne dmontrent pas de fixation de Smad3 sur le promoteur dhTERT (ralises par immunoprcipitation de la chromatine), ni dinteraction entre c-myc et Smad3 (rsultats non montrs). Une tude toute rcente dcrit quant elle un mcanisme dinhibition de la transcription du gne hTERT par le TGF- faisant intervenir la voie des Smads, mais aussi les voies Erk et p38MAPK, avec comme inhibiteur de la transcription la protine E2F-1, qui se fixe sur quatre sites au niveau du promoteur du gne hTERT sans perturber la fixation de c-myc (Lacerte, Korah et al. 2007). Le mode daction prcis du TGFdans la rgulation du gne hTERT dans notre modle hmatologique, reste lucider. Etant donn la rgulation de la tlomrase par deux cytokines de lrythropose impliques dans la prolifration, la survie et la diffrenciation, nous avons mis lhypothse que la tlomrase puisse jouer un rle prpondrant dans lexpansion de la ligne rythrode. Lobtention de cellules UT7 dpltes en tlomrase (UT7-DN-hTERT) grce linfection rtrovirale par un plasmide codant pour le dominant ngatif de la tlomrase, a permis dapporter certains lments de rponse cette question. Les

70

premires expriences ralises montrent que, dans le modle UT7, la tlomrase ne joue aucun rle dans la prolifration induite par lEPO ou la protection contre lapoptose. En revanche, les travaux de Delhommeau et coll. montrent que linhibition de la tlomrase induit un phnotype snescent dans les cellules UT7 au bout de 4 semaines de culture, avec raccourcissement des tlomres et inhibition de la prolifration long terme (Delhommeau, Thierry et al. 2002). De plus, des rsultats prliminaires obtenus lors de notre travail montrent quun ligand des G-quadruplex, le 12459 (obtenu en collaboration avec lquipe du Dr. JF. Riou, IFR53 Reims), dont leffet permet une inhibition de la tlomrase par pissage alternatif (Gomez, Lemarteleur et al. 2004), induit une inhibition de la formation des colonies rythrodes BFU-E, lors de culture de cellules CD34+ en mthylcellulose (Figure 23). Ces rsultats doivent tre rapports aux observations rcentes de Schuller et coll. montrant une corrlation entre la longueur des tlomres et lactivit tlomrase, et les capacits dexpansion rythrode (Schuller, Jankowski et al. 2007), ainsi quaux travaux de Zimmermann et coll. qui suggrent que la tlomrase puisse tre implique dans la diffrenciation rythrode prcoce puisque sa surexpression dans des cellules CD34+, induit une augmentation du nombre de BFU-E obtenues en mthylcellulose (Zimmermann, Glaser et al. 2004). On peut donc mettre lhypothse que toute altration de la rgulation de la tlomrase par lEPO puisse aboutir un appauvrissement de la production des cellules rythrodes, et donc une anmie.

A 5

B
% de formation de BFU-E

120 100 80 60 40 20 0 0 20 40 Dose de 12459 en M 60

IC50 = 24 M

12459

Figure 23 : A) Mode daction du 12459, stabilisation des structures G-quadruplex sur lADN. B) Effet du 12459, un ligand des G-quadruplex, sur la formation des colonies rythrodes (BFU-E) en mthylcellulose.

71

Le modle UT7 ne nous a pas permis dexplorer de manire convaincante, le rle de la tlomrase dans la diffrenciation rythrode. Etant donn les tudes publies ce sujet et les rsultats obtenus dans notre travail, on peut cependant penser que la tlomrase pourrait tre implique dans le processus de diffrenciation. En effet, nous avons pu observer que lors de la phase de diffrenciation terminale des cellules rythrodes, lactivit de la tlomrase diminuait de manire concomitante. Ce rsultat a dailleurs pu tre observ dans plusieurs autres modles dont la ligne mylode (Beyne-Rauzy, PradeHoudellier et al. 2005), les cellules germinales des testicules (Schrader, Burger et al. 2002), les cellules pithliales (Crowe, Nguyen et al. 2005), ou les cellules souches embryonnaires de souris (Wang, Hu et al. 2007). De plus, le fait que le TGF- inhibe la tlomrase tout en acclrant la diffrenciation rythrode est en accord avec lhypothse avance. Afin dapporter des lments de rponse, il pourrait tre intressant de moduler lexpression de la tlomrase par infection rtrovirale de plasmides permettant la surexpression dhTERT ou son inhibition par un dominant ngatif, dans les cellules primaires au stade de progniteurs rythrodes (aprs le tri en CD36+) (Figure 24). Leffet attendu sera alors, une acclration de la diffrenciation des progniteurs rythrodes en labsence de tlomrase catalytiquement active, entranant une rythropose abortive.
CD34+
SCF IL3 IL6 CFU-GEMM

CD36+

CFU-E

Erythroblaste basophile

Cellule souche

BFU-E

+ Epo

Erythroblaste acidophile

Tri des cellules CD36+

Mig-R

Mig-RDN-hTERT

Infection rtrovirale
CD36+
CFU-E

Prolifration ? Survie ?

Mig-RhTERT

+ Epo

Diffrenciation ?

Figure 24 : Effet de la modulation de la tlomrase dans les progniteurs rythrodes humains ?

72

DISCUSSION ET PERSPECTIVES

Ce travail visant dterminer les modes de rgulation de la tlomrase au cours de lhmatopose, est compos de deux thmatiques. La premire concerne ltude de voies de signalisation impliques dans la rgulation de la tlomrase par le TNF- et le GMCSF, dans les cellules mylodes normales et leucmiques. La deuxime a pour objet la caractrisation des signaux permettant la rgulation de la tlomrase par des cytokines de lrythropose, lrythropotine et le TGF- . Lors de cette deuxime partie nous nous sommes de plus intresss la fonction de la tlomrase dans le processus drythropose.

La premire partie de notre travail nous a permis de montrer que dans des cellules de LAM ainsi que dans des progniteurs hmatopotiques normaux, la tlomrase tait inhibe transcriptionnellement par le TNF- , et ce via la production de cramide et lactivation de la voie JNK. Dautre part, le GM-CSF permet de bloquer cette inhibition et donc de restaurer une activit tlomrase normale dans les cellules. Des expriences complmentaires permettront de dterminer les facteurs de transcription spcifiquement impliqus dans ce modle (c-myc, AP-1), ainsi que le mode daction du GM-CSF. Ltude prcdemment mene dans notre laboratoire (Beyne-Rauzy, Recher et al. 2004) a dmontr que cette inhibition de la tlomrase par le TNF- aboutissait lapparition dun phnotype de snescence dans les cellules de LAM et les progniteurs hmatopotiques normaux, et que le GM-CSF permettait de restaurer la prolifration des cellules. Ce phnotype de snescence tant accompagn dune forte instabilit chromosomique, ces donnes trouvent leur importance dans les mcanismes de leucmognse. Le statut dhTERT au cours de la leucmognse reste quelque peu ambigu. En effet, la surexpression de la tlomrase semble tre un vnement important dans lexpansion tumorale puisque 90% des cellules cancreuses prsentent un niveau de tlomrase lev. Cependant, une inhibition chronique de la tlomrase induisant de linstabilit chromosomique lie au phnotype de snescence, peut favoriser lmergence dun clone leucmique en permettant lapparition de mutations ou translocations oncogniques. Ainsi, dans certaines pathologies pr-leucmiques telles que les syndromes mylodysplasiques caractriss par la co-existence dune hmatopose

73

normale et dune hmatopose leucmique, et o lon retrouve un excs de production de TNF- , la rgulation de la tlomrase pourra avoir plusieurs effets. On peut tout dabord penser que le TNF- , en inhibant la tlomrase pourrait soit participer la limitation de lexpansion du clone leucmique, soit au contraire favoriser son mergence en induisant un contexte dinstabilit chromosomique. La dtermination de leffet du TNFsur lvolution physiopathologique de cette maladie, nous semble dautant plus importante que certaines maladies inflammatoires et les syndromes mylodysplasiques peuvent tre traits par des anti-TNF- , dont limpact sur la rgulation de la tlomrase reste inconnu. Les observations ralises dans la deuxime partie de notre travail, suggrent que la tlomrase peut tre rgule transcriptionnellement de faon positive par un facteur de croissance, lEPO, et ngative par une cytokine inhibitrice, le TGF- , au cours de lrythropose normale. Les voies de signalisation respectivement impliques sont, la voie JAK2/STAT/c-myc pour lEPO, et la voie des Smads pour le TGF- . Il est important de prciser que dans notre modle, le TGF- a un effet dominant sur celui de lEPO puisquil permet de bloquer compltement linduction de la transcription du gne hTERT par lEPO lorsquil est ajout dans le milieu de culture en mme temps que lEPO. Contrairement dautres tudes, nos expriences de modulation de lactivit tlomrase dans les cellules UT7 montrent que la tlomrase ne joue pas de rle dans linduction de la prolifration par lEPO, mme si lors de la stimulation par lEPO, lactivit de la tlomrase augmente avant lentre dans le cycle cellulaire (Figure 15 p62). De la mme faon, la tlomrase ne semble pas tre implique dans la protection contre lapoptose. Ce rsultat ntait pas attendu, tant donn le rle protecteur de lEPO dans de nombreux modles cellulaires hmatopotiques ou extra-hmatopotiques (Chappell, Tilbrook et al. 1997; Ghezzi and Brines 2004; Pajonk, Weil et al. 2004), ainsi que la dmonstration que la tlomrase pouvait tre implique dans les mcanismes de survie cellulaire (Akiyama, Yamada et al. 2002; Misawa, Tauchi et al. 2002; Wu, Meng et al. 2005; Li, Ferguson et al. 2006; Gao and Chen 2007). Cependant, les travaux raliss par Delhommeau et coll. montrent que la tlomrase est implique dans lexpansion des cellules hmatopotiques long terme (Delhommeau, Thierry et al. 2002). Dans les cellules hmatopotiques normales, lensemble de nos rsultats dmontre que lactivit tlomrase augmente de faon importante lors de la phase dexpansion des

74

progniteurs, sous laction dun mlange de cytokines, puis diminue lors de la phase de diffrenciation finale tant mylode qurythrode. En parallle, des cytokines inflammatoires telles que le TNF- ou le TGF- , ajouts dans le milieu de culture, inhibent la transcription du gne hTERT et donc son activit, tout en limitant la prolifration et en acclrant la diffrenciation des cellules hmatopotiques. Lensemble de ces observations suggre que la tlomrase est fortement implique dans le processus dhmatopose, puisque tout bouleversement de son niveau dactivit par des signaux extrieurs tels que le TNF- ou le TGF- peut modifier de faon importante le droulement du processus de diffrenciation hmatopotique en aboutissant une hmatopose inefficace. Des expriences de modulation de la tlomrase dans les progniteurs hmatopotiques permettront de vrifier cette hypothse. En effet, on peut supposer que lexpression dune tlomrase constitutivement active dans des progniteurs hmatopotiques, bloquerait les effets dinhibition de la prolifration et dacclration de la diffrenciation induits par le TNFet le TGF- . Et que son inhibition induirait au contraire une hmatopose inefficace.

Nous pouvons proposer deux perspectives principales afin de poursuivre notre travail. 1) Au cours de notre tude, des rsultats prliminaires obtenus tant sur le modle UT7

que KG1 suggrent que la tlomrase peut tre rgule transcriptionnellement par des tyrosines kinases telles que Syk ou les protines de la famille Src (Figure 25 A). Ces protines tant retrouves surexprimes dans certaines leucmies, les lymphomes anaplasiques ALK+ (Thompson, Stumph et al. 2005) ou les lymphomes folliculaires (Leseux, Hamdi et al. 2006) pour Syk, et les LAM pour les Src (Ozawa et coll. Abstract 2766, ASH 2005), ltude de la rgulation de la tlomrase par ces tyrosines kinases nous parait importante. En effet, dans les LAM, une forte augmentation de lactivit de la tlomrase a t retrouve dans plus de 70% des cas, sans que les mcanismes de cette surexpression ne soient lucids. Ainsi, si linhibition cible de ces protines permet de bloquer linduction de lactivit tlomrase dans les leucmies, ce projet peut permettre lidentification de nouvelles cibles thrapeutiques spcifiques des cellules leucmiques (Figure 25 B).

75

A
3

57

transcrit hTERT/PBGD

2,5 2 1,5 1 0,5 0

STAT

5
0 Epo 6h SU6656 6h Picea 6h

'

Figure 25 : Rgulation de la tlomrase par les tyrosines kinases. A) Rsultats prliminaires obtenus dans la ligne UT7. Effet des inhibiteurs de Src (SU6656 10M) et de Syk (Picatanol 20M) sur le nombre de transcrit hTERT (mesur en RT-PCR temps rel) lors dune stimulation des cellules lEPO. B) Schma hypothtique de rgulation de la tlomrase dans les leucmies.

2)

Nous avons vu que la tlomrase tait soumise une somme importante de

rgulation par les facteurs de croissance et cytokines de lrythropose. On peut donc supposer que toute modification de ces voies de rgulation puisse avoir un effet sur le bon droulement du processus de diffrenciation des cellules rouges. Ainsi, dans une pathologie telle que la polyglobulie primitive, la prsence de la mutation activatrice JAK2V617F pourrait induire une suractivation de la tlomrase et expliquer la trop forte expansion de la ligne rythrocytaire. Au contraire, laccumulation de cytokines inflammatoires dans le microenvironnement mdullaire des syndromes mylodysplasiques induirait une inhibition de la tlomrase rendant compte de lrythropose inefficace et donc de lanmie prsente dans 90% des cas (Figure 26). Les tudes ralises sur le statut de la tlomrase dans les mylodysplasies nont pas montr de diffrences majeures de niveau dexpression par rapport aux cellules normales, mais ces rsultats ont t obtenus sur la totalit des cellules mononucles. Nous envisageons pour notre part, dvaluer le niveau dactivit tlomrase ainsi que la longueur des tlomres, dans la population rythrocytaire des chantillons de moelle de patients atteints de SMD. Cette tude pourrait

76

ainsi permettre de mieux comprendre les mcanismes physiopathologiques des anmies retrouves dans les SMD, mais aussi dans les maladies inflammatoires et chez le sujet g.

Polyglobulie primitive : rythropose trop importante

JAK2 V617F

Erythropose normale

Mylodysplasie : rythropose inefficace

TNF- , TGF- , ...

hTERT

hTERT

?
Niveau dactivit tlomrase

Figure 26 : Influence de la rgulation de la tlomrase sur lrythropose dans des pathologies hmatologiques.

77

78

REFERENCES

Adam-Klages, S., D. Adam, et al. (1996). "FAN, a novel WD-repeat protein, couples the p55 TNF-receptor to neutral sphingomyelinase." Cell 86(6): 937-47. Aisner, D. L., W. E. Wright, et al. (2002). "Telomerase regulation: not just flipping the switch." Curr Opin Genet Dev. 12(1): 80-5. Akalin, A., L. W. Elmore, et al. (2001). "A novel mechanism for chaperone-mediated telomerase regulation during prostate cancer progression." Cancer Res. 61(12): 4791-6. Akiyama, M., T. Hideshima, et al. (2002). "Cytokines modulate telomerase activity in a human multiple myeloma cell line." Cancer Res. 62(13): 3876-82. Akiyama, M., T. Hideshima, et al. (2003). "Nuclear factor-kappaB p65 mediates tumor necrosis factor alpha-induced nuclear translocation of telomerase reverse transcriptase protein." Cancer Res. 63(1): 18-21. Akiyama, M., O. Yamada, et al. (2004). "TNFalpha induces rapid activation and nuclear translocation of telomerase in human lymphocytes." Biochem Biophys Res Commun. 316(2): 528-32. Akiyama, M., O. Yamada, et al. (2002). "Telomerase overexpression in K562 leukemia cells protects against apoptosis by serum deprivation and double-stranded DNA break inducing agents, but not against DNA synthesis inhibitors." Cancer Lett 178(2): 187-97. Alfonso-De Matte, M. Y., H. Yang, et al. (2002). "Telomerase is regulated by c-Jun NH2terminal kinase in ovarian surface epithelial cells." Cancer Res 62(16): 4575-8. Allouche, M., A. Bettaieb, et al. (1997). "Influence of Bcl-2 overexpression on the ceramide pathway in daunorubicin-induced apoptosis of leukemic cells." Oncogene 14(15): 1837-45. Armstrong, L., G. Saretzki, et al. (2005). "Overexpression of telomerase confers growth advantage, stress resistance, and enhanced differentiation of ESCs toward the hematopoietic lineage." Stem Cells 23(4): 516-29. Artandi, S. E. and R. A. DePinho (2000). "Mice without telomerase: what can they teach us about human cancer?" Nat Med 6(8): 852-5. Autexier, C. and N. F. Lue (2006). "The structure and function of telomerase reverse transcriptase." Annu Rev Biochem. 75: 493-517. Bachand, F., I. Triki, et al. (2001). "Human telomerase RNA-protein interactions." Nucleic Acids Res 29(16): 3385-93.

79

Backx, B., L. Broeders, et al. (1991). "Positive and negative effects of tumor necrosis factor on colony growth from highly purified normal marrow progenitors." Leukemia 5(1): 66-70. Baumann, P., E. Podell, et al. (2002). "Human Pot1 (protection of telomeres) protein: cytolocalization, gene structure, and alternative splicing." Mol Cell Biol 22(22): 8079-87. Bettaieb, A., M. Record, et al. (1996). "Opposite effects of tumor necrosis factor alpha on the sphingomyelin-ceramide pathway in two myeloid leukemia cell lines: role of transverse sphingomyelin distribution in the plasma membrane." Blood 88(4): 1465-72. Beutler, B. (1990). "The complex regulation and biology of TNF (cachectin)." Crit Rev Oncog 2(1): 9-18. Beyne-Rauzy, O., N. Prade-Houdellier, et al. (2005). "Tumor necrosis factor-alpha inhibits hTERT gene expression in human myeloid normal and leukemic cells." Blood 106(9): 3200-5. Beyne-Rauzy, O., C. Recher, et al. (2004). "Tumor necrosis factor alpha induces senescence and chromosomal instability in human leukemic cells." Oncogene. 23(45): 7507-16. Blackburn, E. H. (2001). "Switching and signaling at the telomere." Cell 106(6): 661-73. Blackburn, E. H., C. W. Greider, et al. (1989). "Recognition and elongation of telomeres by telomerase." Genome 31(2): 553-60. Blasco, M. A., H. W. Lee, et al. (1997). "Mouse models for the study of telomerase." Ciba Found Symp 211: 160-70; discussion 170-6. Bock, O., E. Serinsoz, et al. (2004). "Different expression levels of the telomerase catalytic subunit hTERT in myeloproliferative and myelodysplastic diseases." Leuk Res 28(5): 45760. Bodnar, A. G., M. Ouellette, et al. (1998). "Extension of life-span by introduction of telomerase into normal human cells." Science 279(5349): 349-52. Boggess, J. F., C. Zhou, et al. (2006). "Estrogen-receptor-dependent regulation of telomerase activity in human endometrial cancer cell lines." Gynecol Oncol. 103(2): 41724. Epub 2006 May 11. Broccoli, D., A. Smogorzewska, et al. (1997). "Human telomeres contain two distinct Myb-related proteins, TRF1 and TRF2." Nat Genet 17(2): 231-5. Broccoli, D., J. W. Young, et al. (1995). "Telomerase activity in normal and malignant hematopoietic cells." Proc Natl Acad Sci U S A 92(20): 9082-6. Brousset, P., N. Chaouche, et al. (1998). "Telomerase activity in Hodgkin' disease." Leuk s Lymphoma 30(1-2): 189-92.

80

Bryan, T. M., A. Englezou, et al. (1997). "Evidence for an alternative mechanism for maintaining telomere length in human tumors and tumor-derived cell lines." Nat Med 3(11): 1271-4. Bryan, T. M., A. Englezou, et al. (1995). "Telomere elongation in immortal human cells without detectable telomerase activity." Embo J 14(17): 4240-8. Budiyanto, A., T. Bito, et al. (2003). "Inhibition of the epidermal growth factor receptor suppresses telomerase activity in HSC-1 human cutaneous squamous cell carcinoma cells." J Invest Dermatol. 121(5): 1088-94. Campisi, J., S. H. Kim, et al. (2001). "Cellular senescence, cancer and aging: the telomere connection." Exp Gerontol 36(10): 1619-37. Cao, Y., H. Li, et al. (2002). "TERT regulates cell survival independent of telomerase enzymatic activity." Oncogene 21(20): 3130-8. Caux, C., C. Favre, et al. (1991). "Potentiation of early hematopoiesis by tumor necrosis factor-alpha is followed by inhibition of granulopoietic differentiation and proliferation." Blood 78(3): 635-44. Cerezo, A., H. J. Stark, et al. (2003). "Constitutive overexpression of human telomerase reverse transcriptase but not c-myc blocks terminal differentiation in human HaCaT skin keratinocytes." J Invest Dermatol 121(1): 110-9. Chai, W., L. P. Ford, et al. (2002). "Human Ku70/80 associates physically with telomerase through interaction with hTERT." J Biol Chem 277(49): 47242-7. Chan, S. W. and E. H. Blackburn (2002). "New ways not to make ends meet: telomerase, DNA damage proteins and heterochromatin." Oncogene 21(4): 553-63. Chappell, D., P. A. Tilbrook, et al. (1997). "Prevention of apoptosis in J2E erythroid cells by erythropoietin: involvement of JAK2 but not MAP kinases." Cell Death Differ 4(2): 105-13. Chen, C. and A. J. Sytkowski (2001). "Erythropoietin activates two distinct signaling pathways required for the initiation and the elongation of c-myc." J Biol Chem 276(42): 38518-26. Chen, J. L., M. A. Blasco, et al. (2000). "Secondary structure of vertebrate telomerase RNA." Cell. 100(5): 503-14. Chiu, C. P., W. Dragowska, et al. (1996). "Differential expression of telomerase activity in hematopoietic progenitors from adult human bone marrow." Stem Cells 14(2): 239-48. Cohen, S. B., M. E. Graham, et al. (2007). "Protein composition of catalytically active human telomerase from immortal cells." Science 315(5820): 1850-3. Colgin, L. M., C. Wilkinson, et al. (2000). "The hTERTalpha splice variant is a dominant negative inhibitor of telomerase activity." Neoplasia. 2(5): 426-32.

81

Cong, Y. S. and S. Bacchetti (2000). "Histone deacetylation is involved in the transcriptional repression of hTERT in normal human cells." J Biol Chem. 275(46): 356658. Cong, Y. S., J. Wen, et al. (1999). "The human telomerase catalytic subunit hTERT: organization of the gene and characterization of the promoter." Hum Mol Genet. 8(1): 13742. Conneally, E., J. Cashman, et al. (1997). "Expansion in vitro of transplantable human cord blood stem cells demonstrated using a quantitative assay of their lympho-myeloid repopulating activity in nonobese diabetic-scid/scid mice." Proc Natl Acad Sci U S A 94(18): 9836-41. Crowe, D. L., D. C. Nguyen, et al. (2005). "Mechanism of telomerase repression during terminal differentiation of normal epithelial cells and squamous carcinoma lines." Int J Oncol 27(3): 847-54. Crowe, D. L., D. C. Nguyen, et al. (2001). "E2F-1 represses transcription of the human telomerase reverse transcriptase gene." Nucleic Acids Res. 29(13): 2789-94. De Maria, R., A. Zeuner, et al. (1999). "Negative regulation of erythropoiesis by caspasemediated cleavage of GATA-1." Nature 401(6752): 489-93. Delhommeau, F., A. Thierry, et al. (2002). "Telomere dysfunction and telomerase reactivation in human leukemia cell lines after telomerase inhibition by the expression of a dominant-negative hTERT mutant." Oncogene 21(54): 8262-71. Dhar, S., J. A. Squire, et al. (2000). "Inactivation of 14-3-3sigma influences telomere behavior and ionizing radiation-induced chromosomal instability." Mol Cell Biol. 20(20): 7764-72. Dimri, G. P., X. Lee, et al. (1995). "A biomarker that identifies senescent human cells in culture and in aging skin in vivo." Proc Natl Acad Sci U S A 92(20): 9363-7. Dunham, M. A., A. A. Neumann, et al. (2000). "Telomere maintenance by recombination in human cells." Nat Genet 26(4): 447-50. Dybedal, I., D. Bryder, et al. (2001). "Tumor necrosis factor (TNF)-mediated activation of the p55 TNF receptor negatively regulates maintenance of cycling reconstituting human hematopoietic stem cells." Blood 98(6): 1782-91. Engelhardt, M., P. Drullinsky, et al. (1997). "Telomerase and telomere length in the development and progression of premalignant lesions to colorectal cancer." Clin Cancer Res 3(11): 1931-41. Engelhardt, M., R. Kumar, et al. (1997). "Telomerase regulation, cell cycle, and telomere stability in primitive hematopoietic cells." Blood 90(1): 182-93. Engelhardt, M., K. Mackenzie, et al. (2000). "Telomerase activity and telomere length in acute and chronic leukemia, pre- and post-ex vivo culture." Cancer Res 60(3): 610-7.

82

Engelhardt, M., M. F. Ozkaynak, et al. (1998). "Telomerase activity and telomere length in pediatric patients with malignancies undergoing chemotherapy." Leukemia 12(1): 13-24. Engelhardt, M., R. Wasch, et al. (2004). "Telomeres and telomerase in normal and leukemic hematopoietic cells." Leuk Res 28(10): 1001-4. Ferraris, A. M., R. Mangerini, et al. (2005). "High telomerase activity in granulocytes from clonal polycythemia vera and essential thrombocythemia." Blood 105(5): 2138-40. Ford, L. P., J. W. Shay, et al. (2001). "The La antigen associates with the human telomerase ribonucleoprotein and influences telomere length in vivo." Rna. 7(8): 1068-75. Ford, L. P., J. M. Suh, et al. (2000). "Heterogeneous nuclear ribonucleoproteins C1 and C2 associate with the RNA component of human telomerase." Mol Cell Biol. 20(23): 9084-91. Forsythe, H. L., L. W. Elmore, et al. (2002). "Retroviral-mediated expression of telomerase in normal human cells provides a selective growth advantage." Int J Oncol 20(6): 1137-43. Forsythe, H. L., J. L. Jarvis, et al. (2001). "Stable association of hsp90 and p23, but Not hsp70, with active human telomerase." J Biol Chem. 276(19): 15571-4. Epub 2001 Mar 23. Franco, S., M. F. Ozkaynak, et al. (2003). "Telomere dynamics in childhood leukemia and solid tumors: a follow-up study." Leukemia 17(2): 401-10. Freyssinier, J. M., C. Lecoq-Lafon, et al. (1999). "Purification, amplification and characterization of a population of human erythroid progenitors." Br J Haematol 106(4): 912-22. Fujimoto, K., S. Kyo, et al. (2000). "Identification and characterization of negative regulatory elements of the human telomerase catalytic subunit (hTERT) gene promoter: possible role of MZF-2 in transcriptional repression of hTERT." Nucleic Acids Res. 28(13): 2557-62. Gandellini, P., M. Folini, et al. (2007). "Down-regulation of human telomerase reverse transcriptase through specific activation of RNAi pathway quickly results in cancer cell growth impairment." Biochem Pharmacol 73(11): 1703-14. Gao, X. D. and Y. R. Chen (2007). "Inhibition of telomerase with human telomerase reverse transcriptase antisense enhances tumor necrosis factor-alpha-induced apoptosis in bladder cancer cells." Chin Med J (Engl) 120(9): 755-60. Ghezzi, P. and M. Brines (2004). "Erythropoietin as an antiapoptotic, tissue-protective cytokine." Cell Death Differ 11 Suppl 1: S37-44. Gire, V. and D. Wynford-Thomas (1998). "Reinitiation of DNA synthesis and cell division in senescent human fibroblasts by microinjection of anti-p53 antibodies." Mol Cell Biol 18(3): 1611-21.

83

Gisselsson, D., T. Jonson, et al. (2001). "Telomere dysfunction triggers extensive DNA fragmentation and evolution of complex chromosome abnormalities in human malignant tumors." Proc Natl Acad Sci U S A 98(22): 12683-8. Gomez, D., T. Lemarteleur, et al. (2004). "Telomerase downregulation induced by the Gquadruplex ligand 12459 in A549 cells is mediated by hTERT RNA alternative splicing." Nucleic Acids Res 32(1): 371-9. Gorbunova, V., A. Seluanov, et al. (2002). "Expression of human telomerase (hTERT) does not prevent stress-induced senescence in normal human fibroblasts but protects the cells from stress-induced apoptosis and necrosis." J Biol Chem 277(41): 38540-9. Goueli, B. S. and R. Janknecht (2004). "Upregulation of the Catalytic Telomerase Subunit by the Transcription Factor ER81 and Oncogenic HER2/Neu, Ras, or Raf." Mol Cell Biol 24(1): 25-35. Goueli, B. S. and R. Janknecht (2004). "Upregulation of the Catalytic Telomerase Subunit by the Transcription Factor ER81 and Oncogenic HER2/Neu, Ras, or Raf." Mol Cell Biol. 24(1): 25-35. Greenberg, R. A., R. C. O' Hagan, et al. (1999). "Telomerase reverse transcriptase gene is a direct target of c-Myc but is not functionally equivalent in cellular transformation." Oncogene. 18(5): 1219-26. Greider, C. W. and E. H. Blackburn (1985). "Identification of a specific telomere terminal transferase activity in Tetrahymena extracts." Cell. 43(2 Pt 1): 405-13. Greider, C. W. and E. H. Blackburn (1989). "A telomeric sequence in the RNA of Tetrahymena telomerase required for telomere repeat synthesis." Nature. 337(6205): 3317. Griffith, J. D., L. Comeau, et al. (1999). "Mammalian telomeres end in a large duplex loop." Cell 97(4): 503-14. Gunes, C., S. Lichtsteiner, et al. (2000). "Expression of the hTERT gene is regulated at the level of transcriptional initiation and repressed by Mad1." Cancer Res. 60(8): 2116-21. Haendeler, J., J. Hoffmann, et al. (2003). "Hydrogen peroxide triggers nuclear export of telomerase reverse transcriptase via Src kinase family-dependent phosphorylation of tyrosine 707." Mol Cell Biol 23(13): 4598-610. Hande, M. P., E. Samper, et al. (1999). "Telomere length dynamics and chromosomal instability in cells derived from telomerase null mice." J Cell Biol 144(4): 589-601. Hayflick, L. (1965). "The Limited In Vitro Lifetime Of Human Diploid Cell Strains." Exp Cell Res 37: 614-36. Herrera, E., A. C. Martinez, et al. (2000). "Impaired germinal center reaction in mice with short telomeres." Embo J 19(3): 472-81.

84

Hiyama, K., Y. Hirai, et al. (1995). "Activation of telomerase in human lymphocytes and hematopoietic progenitor cells." J Immunol 155(8): 3711-5. Hoare, S. F., L. A. Bryce, et al. (2001). "Lack of telomerase RNA gene hTERC expression in alternative lengthening of telomeres cells is associated with methylation of the hTERC promoter." Cancer Res. 61(1): 27-32. Hou, M., D. Xu, et al. (2001). "Real-time quantitative telomeric repeat amplification protocol assay for the detection of telomerase activity." Clin Chem 47(3): 519-24. Hu, B., D. C. Tack, et al. (2006). "Role of Smad3 in the regulation of rat telomerase reverse transcriptase by TGFbeta." Oncogene 25(7): 1030-41. Iczkowski, K. A., W. Huang, et al. (2004). "Androgen ablation therapy for prostate carcinoma suppresses the immunoreactive telomerase subunit hTERT." Cancer 100(2): 294-9. Igarashi, H. and N. Sakaguchi (1997). "Telomerase activity is induced in human peripheral B lymphocytes by the stimulation to antigen receptor." Blood 89(4): 1299-307. Jacobsen, F. W., M. Rothe, et al. (1994). "Role of the 75-kDa tumor necrosis factor receptor: inhibition of early hematopoiesis." Proc Natl Acad Sci U S A 91(22): 10695-9. James, C., V. Ugo, et al. (2005). "A unique clonal JAK2 mutation leading to constitutive signalling causes polycythaemia vera." Nature 434(7037): 1144-8. Johnson, R. A., T. A. Waddelow, et al. (1989). "Chronic exposure to tumor necrosis factor in vivo preferentially inhibits erythropoiesis in nude mice." Blood 74(1): 130-8. Jun, E. S., T. H. Lee, et al. (2004). "Expression of telomerase extends longevity and enhances differentiation in human adipose tissue-derived stromal cells." Cell Physiol Biochem 14(4-6): 261-8. Kaminker, P. G., S. H. Kim, et al. (2001). "TANK2, a new TRF1-associated poly(ADPribose) polymerase, causes rapid induction of cell death upon overexpression." J Biol Chem 276(38): 35891-9. Kanaya, T., S. Kyo, et al. (2000). "Adenoviral expression of p53 represses telomerase activity through down-regulation of human telomerase reverse transcriptase transcription." Clin Cancer Res 6(4): 1239-47. Kang, H. J., Y. S. Choi, et al. (2004). "Ectopic expression of the catalytic subunit of telomerase protects against brain injury resulting from ischemia and NMDA-induced neurotoxicity." J Neurosci 24(6): 1280-7. Kang, S. S., T. Kwon, et al. (1999). "Akt protein kinase enhances human telomerase activity through phosphorylation of telomerase reverse transcriptase subunit." J Biol Chem. 274(19): 13085-90.

85

Kelly, M. L., Y. Tang, et al. (1998). "Granulocyte-macrophage colony-stimulating factor rescues TF-1 leukemia cells from ionizing radiation-induced apoptosis through a pathway mediated by protein kinase Calpha." Blood 92(2): 416-24. Keppler, B. R. and M. B. Jarstfer (2004). "Inhibition of telomerase activity by preventing proper assemblage." Biochemistry. 43(2): 334-43. Kharbanda, S., V. Kumar, et al. (2000). "Regulation of the hTERT telomerase catalytic subunit by the c-Abl tyrosine kinase." Curr Biol. 10(10): 568-75. Kim, N. W., M. A. Piatyszek, et al. (1994). "Specific association of human telomerase activity with immortal cells and cancer." Science 266(5193): 2011-5. Kim, S. H., P. Kaminker, et al. (1999). "TIN2, a new regulator of telomere length in human cells." Nat Genet 23(4): 405-12. Kimura, A., M. Ohmichi, et al. (2004). "Induction of hTERT expression and phosphorylation by estrogen via Akt cascade in human ovarian cancer cell lines." Oncogene. 23(26): 4505-15. Kirkpatrick, K. L., G. Clark, et al. (2003). "hTERT mRNA expression correlates with telomerase activity in human breast cancer." Eur J Surg Oncol. 29(4): 321-6. Kolquist, K. A., L. W. Ellisen, et al. (1998). "Expression of TERT in early premalignant lesions and a subset of cells in normal tissues." Nat Genet 19(2): 182-6. Koyanagi, Y., D. Kobayashi, et al. (2000). "Telomerase activity is down regulated via decreases in hTERT mRNA but not TEP1 mRNA or hTERC during the differentiation of leukemic cells." Anticancer Res 20(2A): 773-8. Ku, W. C., A. J. Cheng, et al. (1997). "Inhibition of telomerase activity by PKC inhibitors in human nasopharyngeal cancer cells in culture." Biochem Biophys Res Commun. 241(3): 730-6. Kurz, D. J., Y. Hong, et al. (2003). "Fibroblast growth factor-2, but not vascular endothelial growth factor, upregulates telomerase activity in human endothelial cells." Arterioscler Thromb Vasc Biol. 23(5): 748-54. Epub 2003 Apr 3. Kyo, S., M. Takakura, et al. (1999). "Estrogen activates telomerase." Cancer Res. 59(23): 5917-21. Kyo, S., M. Takakura, et al. (2000). "Sp1 cooperates with c-Myc to activate transcription of the human telomerase reverse transcriptase gene (hTERT)." Nucleic Acids Res. 28(3): 669-77. Lacerte, A., J. Korah, et al. (2007). "Transforming growth factor-beta inhibits telomerase through SMAD3 and E2F transcription factors." Cell Signal.

86

Lansdorp, P. M., H. J. Sutherland, et al. (1990). "Selective expression of CD45 isoforms on functional subpopulations of CD34+ hemopoietic cells from human bone marrow." J Exp Med 172(1): 363-6. Le, S., R. Sternglanz, et al. (2000). "Identification of two RNA-binding proteins associated with human telomerase RNA." Mol Biol Cell. 11(3): 999-1010. Lee, H. W., M. A. Blasco, et al. (1998). "Essential role of mouse telomerase in highly proliferative organs." Nature 392(6676): 569-74. Lee, J. J., H. Kook, et al. (2001). "Telomere length changes in patients with aplastic anaemia." Br J Haematol 112(4): 1025-30. Lee, S. H., J. W. Kim, et al. (2003). "Interferon regulatory factor-1 (IRF-1) is a mediator for interferon-gamma induced attenuation of telomerase activity and human telomerase reverse transcriptase (hTERT) expression." Oncogene. 22(3): 381-91. Lee, S. H., J. W. Kim, et al. (2005). "IFN-gamma/IRF-1-induced p27kip1 down-regulates telomerase activity and human telomerase reverse transcriptase expression in human cervical cancer." FEBS Lett. 579(5): 1027-33. Epub 2005 Jan 13. Lendvay, T. S., D. K. Morris, et al. (1996). "Senescence mutants of Saccharomyces cerevisiae with a defect in telomere replication identify three additional EST genes." Genetics. 144(4): 1399-412. Leseux, L., S. M. Hamdi, et al. (2006). "Syk-dependent mTOR activation in follicular lymphoma cells." Blood 108(13): 4156-62. Li, B., S. Oestreich, et al. (2000). "Identification of human Rap1: implications for telomere evolution." Cell 101(5): 471-83. Li, H., D. Xu, et al. (2006). "Transforming growth factor beta suppresses human telomerase reverse transcriptase (hTERT) by Smad3 interactions with c-Myc and the hTERT gene." J Biol Chem. 281(35): 25588-600. Epub 2006 Jun 19. Li, H., L. Zhao, et al. (1998). "Telomerase is controlled by protein kinase Calpha in human breast cancer cells." J Biol Chem. 273(50): 33436-42. Li, H., L. L. Zhao, et al. (1997). "Protein phosphatase 2A inhibits nuclear telomerase activity in human breast cancer cells." J Biol Chem. 272(27): 16729-32. Li, S., J. Crothers, et al. (2005). "Cellular and gene expression responses involved in the rapid growth inhibition of human cancer cells by RNA interference-mediated depletion of telomerase RNA." J Biol Chem 280(25): 23709-17. Li, S., M. J. Ferguson, et al. (2006). "Human telomerase reverse transcriptase protects hematopoietic progenitor TF-1 cells from death and quiescence induced by cytokine withdrawal." Leukemia 20(7): 1270-8.

87

Lin, S. Y. and S. J. Elledge (2003). "Multiple tumor suppressor pathways negatively regulate telomerase." Cell. 113(7): 881-9. Lingner, J. and T. R. Cech (1996). "Purification of telomerase from Euplotes aediculatus: requirement of a primer 3' overhang." Proc Natl Acad Sci U S A. 93(20): 10712-7. Lingner, J., T. R. Hughes, et al. (1997). "Reverse transcriptase motifs in the catalytic subunit of telomerase." Science. 276(5312): 561-7. Liu, K., R. J. Hodes, et al. (2001). "Cutting edge: telomerase activation in human T lymphocytes does not require increase in telomerase reverse transcriptase (hTERT) protein but is associated with hTERT phosphorylation and nuclear translocation." J Immunol. 166(8): 4826-30. Liu, L., J. B. Berletch, et al. (2004). "Telomerase inhibition by retinoids precedes cytodifferentiation of leukemia cells and may contribute to terminal differentiation." Mol Cancer Ther 3(8): 1003-9. Lopatina, N. G., J. C. Poole, et al. (2003). "Control mechanisms in the regulation of telomerase reverse transcriptase expression in differentiating human teratocarcinoma cells." Biochem Biophys Res Commun 306(3): 650-9. Lord, J. D., B. C. McIntosh, et al. (2000). "The IL-2 receptor promotes lymphocyte proliferation and induction of the c-myc, bcl-2, and bcl-x genes through the transactivation domain of Stat5." J Immunol 164(5): 2533-41. Ly, H., E. H. Blackburn, et al. (2003). "Comprehensive structure-function analysis of the core domain of human telomerase RNA." Mol Cell Biol. 23(19): 6849-56. Maida, Y., S. Kyo, et al. (2002). "Direct activation of telomerase by EGF through Etsmediated transactivation of TERT via MAP kinase signaling pathway." Oncogene. 21(26): 4071-9. Mansat-de Mas, V., C. Bezombes, et al. (1999). "Implication of radical oxygen species in ceramide generation, c-Jun N-terminal kinase activation and apoptosis induced by daunorubicin." Mol Pharmacol 56(5): 867-74. Mansat-De Mas, V. B.-R., O. Laurent, G. (2003). "Snescence prmature induite par le stress: implication en pharmacologie anti-tumorale." Hmatologie 9: 161-168. Mansat, V., G. Laurent, et al. (1997). "The protein kinase C activators phorbol esters and phosphatidylserine inhibit neutral sphingomyelinase activation, ceramide generation, and apoptosis triggered by daunorubicin." Cancer Res 57(23): 5300-4. Massard, C., Y. Zermati, et al. (2006). "hTERT: a novel endogenous inhibitor of the mitochondrial cell death pathway." Oncogene 25(33): 4505-14. Masutomi, K., R. Possemato, et al. (2005). "The telomerase reverse transcriptase regulates chromatin state and DNA damage responses." Proc Natl Acad Sci U S A 102(23): 8222-7.

88

McClintock, B. (1941). "The Stability of Broken Ends of Chromosomes in Zea Mays." Genetics. 26(2): 234-82. Means, R. T., Jr. and S. B. Krantz (1993). "Inhibition of human erythroid colony-forming units by tumor necrosis factor requires beta interferon." J Clin Invest 91(2): 416-9. Meeker, A. K., H. J. Sommerfeld, et al. (1996). "Telomerase is activated in the prostate and seminal vesicles of the castrated rat." Endocrinology. 137(12): 5743-6. Micheau, O. and J. Tschopp (2003). "Induction of TNF receptor I-mediated apoptosis via two sequential signaling complexes." Cell 114(2): 181-90. Michishita, E., K. Nakabayashi, et al. (1998). "DNA topoisomerase inhibitors induce reversible senescence in normal human fibroblasts." Biochem Biophys Res Commun 253(3): 667-71. Misawa, M., T. Tauchi, et al. (2002). "Inhibition of human telomerase enhances the effect of chemotherapeutic agents in lung cancer cells." Int J Oncol 21(5): 1087-92. Misiti, S., S. Nanni, et al. (2000). "Induction of hTERT expression and telomerase activity by estrogens in human ovary epithelium cells." Mol Cell Biol. 20(11): 3764-71. Mitchell, J. R., J. Cheng, et al. (1999). "A box H/ACA small nucleolar RNA-like domain at the human telomerase RNA 3' end." Mol Cell Biol. 19(1): 567-76. Mitchell, J. R., E. Wood, et al. (1999). "A telomerase component is defective in the human disease dyskeratosis congenita." Nature 402(6761): 551-5. Morales, A., H. Lee, et al. (2007). "Sphingolipids and cell death". Apoptosis. 12 : 923-939. Mufti, G. J. (2004). "Pathobiology, classification, and diagnosis of myelodysplastic syndrome." Best Pract Res Clin Haematol 17(4): 543-57. Mller (1940). "An analysis of the process of structural change in chromosomes of Drosophilia." J Genet 40: 1-66. Nakamura, M., K. Masutomi, et al. (2005). "Efficient inhibition of human telomerase reverse transcriptase expression by RNA interference sensitizes cancer cells to ionizing radiation and chemotherapy." Hum Gene Ther 16(7): 859-68. Nanni, S., M. Narducci, et al. (2002). "Signaling through estrogen receptors modulates telomerase activity in human prostate cancer." J Clin Invest. 110(2): 219-27. Nugent, C. I. and V. Lundblad (1998). "The telomerase reverse transcriptase: components and regulation." Genes Dev. 12(8): 1073-85. O' Sullivan, J. N., M. P. Bronner, et al. (2002). "Chromosomal instability in ulcerative colitis is related to telomere shortening." Nat Genet 32(2): 280-4.

89

Obeid, L. M., C. M. Linardic, et al. (1993). "Programmed cell death induced by ceramide." Science 259(5102): 1769-71. Ogretmen, B., J. M. Kraveka, et al. (2001). "Molecular mechanisms of ceramide-mediated telomerase inhibition in the A549 human lung adenocarcinoma cell line." J Biol Chem 276(35): 32506-14. Ogretmen, B., D. Schady, et al. (2001). "Role of ceramide in mediating the inhibition of telomerase activity in A549 human lung adenocarcinoma cells." J Biol Chem 276(27): 24901-10. Oh, S., Y. Song, et al. (1999). "The Wilms' tumor 1 tumor suppressor gene represses transcription of the human telomerase reverse transcriptase gene." J Biol Chem. 274(52): 37473-8. Oh, S., Y. H. Song, et al. (2000). "Identification of Mad as a repressor of the human telomerase (hTERT) gene." Oncogene. 19(11): 1485-90. Ohyashiki, J. H., K. Ohyashiki, et al. (1997). "Clinical implications of telomerase activity levels in acute leukemia." Clin Cancer Res 3(4): 619-25. Ohyashiki, J. H., G. Sashida, et al. (2002). "Telomeres and telomerase in hematologic neoplasia." Oncogene 21(4): 680-7. Ohyashiki, K., J. H. Ohyashiki, et al. (1997). "Telomerase activity and cytogenetic changes in chronic myeloid leukemia with disease progression." Leukemia 11(2): 190-4. Olovnikov, A. M. (1973). "A theory of marginotomy. The incomplete copying of template margin in enzymic synthesis of polynucleotides and biological significance of the phenomenon." J Theor Biol. 41(1): 181-90. Pajonk, F., A. Weil, et al. (2004). "The erythropoietin-receptor pathway modulates survival of cancer cells." Oncogene 23(55): 8987-91. Pirzio, L. M., M. A. Freulet-Marriere, et al. (2004). "Human fibroblasts expressing hTERT show remarkable karyotype stability even after exposure to ionizing radiation." Cytogenet Genome Res 104(1-4): 87-94. Poon, S. S., U. M. Martens, et al. (1999). "Telomere length measurements using digital fluorescence microscopy." Cytometry 36(4): 267-78. Prowse, K. R. and C. W. Greider (1995). "Developmental and tissue-specific regulation of mouse telomerase and telomere length." Proc Natl Acad Sci U S A 92(11): 4818-22. Quillet-Mary, A., V. Mansat, et al. (1996). "Daunorubicin-induced internucleosomal DNA fragmentation in acute myeloid cell lines." Leukemia 10(3): 417-25. Rahman, R., L. Latonen, et al. (2005). "hTERT antagonizes p53-induced apoptosis independently of telomerase activity." Oncogene 24(8): 1320-7.

90

Rama, S., Y. Suresh, et al. (2001). "Regulation of telomerase during human placental differentiation: a role for TGFbeta1." Mol Cell Endocrinol 182(2): 233-48. Rama, S., Y. Suresh, et al. (2003). "TGF beta1 induces multiple independent signals to regulate human trophoblastic differentiation: mechanistic insights." Mol Cell Endocrinol 206(1-2): 123-36. Ravindranath, N., S. L. Ioffe, et al. (2001). "Androgen depletion activates telomerase in the prostate of the nonhuman primate, Macaca mulatta." Prostate. 49(1): 79-89. Rebel, V. I., S. Hartnett, et al. (1999). "Essential role for the p55 tumor necrosis factor receptor in regulating hematopoiesis at a stem cell level." J Exp Med 190(10): 1493-504. Reddel, R. R. (2003). "Alternative lengthening of telomeres, telomerase, and cancer." Cancer Lett 194(2): 155-62. Ren, J. G., H. L. Xia, et al. (2001). "Hydroxyl radical-induced apoptosis in human tumor cells is associated with telomere shortening but not telomerase inhibition and caspase activation." FEBS Lett 488(3): 123-32. Richardson, R. M., B. Nguyen, et al. (2007). "Ectopic telomerase expression inhibits neuronal differentiation of NT2 neural progenitor cells." Neurosci Lett 421(2): 168-72. Richmond, T. D., M. Chohan, et al. (2005). "Turning cells red: signal transduction mediated by erythropoietin." Trends Cell Biol 15(3): 146-55. Robertson, D. M., L. Li, et al. (2005). "Characterization of growth and differentiation in a telomerase-immortalized human corneal epithelial cell line." Invest Ophthalmol Vis Sci 46(2): 470-8. Robles, S. J. and G. R. Adami (1998). "Agents that cause DNA double strand breaks lead to p16INK4a enrichment and the premature senescence of normal fibroblasts." Oncogene 16(9): 1113-23. Rogers, J. A. and J. W. Berman (1994). "TNF-alpha inhibits the further development of committed progenitors while stimulating multipotential progenitors in mouse long-term bone marrow cultures." J Immunol 153(10): 4694-703. Rohde, V., H. P. Sattler, et al. (2000). "Expression of the human telomerase reverse transcriptase is not related to telomerase activity in normal and malignant renal tissue." Clin Cancer Res. 6(12): 4803-9. Rudolph, K. L., S. Chang, et al. (1999). "Longevity, stress response, and cancer in aging telomerase-deficient mice." Cell 96(5): 701-12. Rudolph, K. L., M. Millard, et al. (2001). "Telomere dysfunction and evolution of intestinal carcinoma in mice and humans." Nat Genet 28(2): 155-9. Rufer, N., W. Dragowska, et al. (1998). "Telomere length dynamics in human lymphocyte subpopulations measured by flow cytometry." Nat Biotechnol 16(8): 743-7.

91

Rusten, L. S., F. W. Jacobsen, et al. (1994). "Bifunctional effects of tumor necrosis factor alpha (TNF alpha) on the growth of mature and primitive human hematopoietic progenitor cells: involvement of p55 and p75 TNF receptors." Blood 83(11): 3152-9. Saeboe-Larssen, S., E. Fossberg, et al. (2006). "Characterization of novel alternative splicing sites in human telomerase reverse transcriptase (hTERT): analysis of expression and mutual correlation in mRNA isoforms from normal and tumour tissues." BMC Mol Biol. 7: 26. Sakabe, H., N. Yahata, et al. (1998). "Human cord blood-derived primitive progenitors are enriched in CD34+c-kit- cells: correlation between long-term culture-initiating cells and telomerase expression." Leukemia 12(5): 728-34. Samper, E., J. M. Flores, et al. (2001). "Restoration of telomerase activity rescues chromosomal instability and premature aging in Terc-/- mice with short telomeres." EMBO Rep 2(9): 800-7. Schmitt, C. A., J. S. Fridman, et al. (2002). "A senescence program controlled by p53 and p16INK4a contributes to the outcome of cancer therapy." Cell 109(3): 335-46. Schrader, M., A. M. Burger, et al. (2002). "The differentiation status of primary gonadal germ cell tumors correlates inversely with telomerase activity and the expression level of the gene encoding the catalytic subunit of telomerase." BMC Cancer 2: 32. Schuller, C. E., K. Jankowski, et al. (2007). "Telomere length of cord blood-derived CD34(+) progenitors predicts erythroid proliferative potential." Leukemia 21(5): 983-91. Seimiya, H., H. Sawada, et al. (2000). "Involvement of 14-3-3 proteins in nuclear localization of telomerase." Embo J. 19(11): 2652-61. Serakinci, N., R. Christensen, et al. (2007). "Ectopically hTERT expressing adult human mesenchymal stem cells are less radiosensitive than their telomerase negative counterpart." Exp Cell Res 313(5): 1056-67. Sharma, G. G., A. Gupta, et al. (2003). "hTERT associates with human telomeres and enhances genomic stability and DNA repair." Oncogene 22(1): 131-46. Shay, J. W., H. Werbin, et al. (1996). "Telomeres and telomerase in human leukemias." Leukemia 10(8): 1255-61. Shin, K. H., M. K. Kang, et al. (2004). "Introduction of human telomerase reverse transcriptase to normal human fibroblasts enhances DNA repair capacity." Clin Cancer Res 10(7): 2551-60. Sieglova, Z., S. Zilovcova, et al. (2004). "Dynamics of telomere erosion and its association with genome instability in myelodysplastic syndromes (MDS) and acute myelogenous leukemia arising from MDS: a marker of disease prognosis?" Leuk Res 28(10): 1013-21.

92

Sinha-Datta, U., I. Horikawa, et al. (2004). "Transcriptional activation of hTERT through the NF-kappaB pathway in HTLV-I-transformed cells." Blood. 104(8): 2523-31. Epub 2004 Jun 29. Smith, L. L., H. A. Coller, et al. (2003). "Telomerase modulates expression of growthcontrolling genes and enhances cell proliferation." Nat Cell Biol 5(5): 474-9. Song, K., D. Jung, et al. (2000). "Interaction of human Ku70 with TRF2." FEBS Lett 481(1): 81-5. Sutherland, H. J., D. E. Hogge, et al. (1996). "Characterization, quantitation and mobilization of early hematopoietic progenitors: implications for transplantation." Bone Marrow Transplant 18 Suppl 1: S1-4. Takakura, M., S. Kyo, et al. (2005). "Function of AP-1 in transcription of the telomerase reverse transcriptase gene (TERT) in human and mouse cells." Mol Cell Biol 25(18): 8037-43. Takakura, M., S. Kyo, et al. (2001). "Telomerase activation by histone deacetylase inhibitor in normal cells." Nucleic Acids Res. 29(14): 3006-11. Tatematsu, K., J. Nakayama, et al. (1996). "A novel quantitative ' stretch PCR assay'that , detects a dramatic increase in telomerase activity during the progression of myeloid leukemias." Oncogene 13(10): 2265-74. ten Dijke, P. and C. S. Hill (2004). "New insights into TGF-beta-Smad signalling." Trends Biochem Sci 29(5): 265-73. Theimer, C. A., C. A. Blois, et al. (2005). "Structure of the human telomerase RNA pseudoknot reveals conserved tertiary interactions essential for function." Mol Cell. 17(5): 671-82. Thompson, M. A., J. Stumph, et al. (2005). "Differential gene expression in anaplastic lymphoma kinase-positive and anaplastic lymphoma kinase-negative anaplastic large cell lymphomas." Hum Pathol 36(5): 494-504. Uehara, H., G. Nardone, et al. (1999). "Detection of telomerase activity utilizing energy transfer primers: comparison with gel- and ELISA-based detection." Biotechniques 26(3): 552-8. Ulaner, G. A., J. F. Hu, et al. (1998). "Telomerase activity in human development is regulated by human telomerase reverse transcriptase (hTERT) transcription and by alternate splicing of hTERT transcripts." Cancer Res. 58(18): 4168-72. van Horssen, R., T. L. Ten Hagen, et al. (2006). "TNF-alpha in cancer treatment: molecular insights, antitumor effects, and clinical utility." Oncologist 11(4): 397-408. Vaziri, H. and S. Benchimol (1998). "Reconstitution of telomerase activity in normal human cells leads to elongation of telomeres and extended replicative life span." Curr Biol 8(5): 279-82.

93

Vaziri, H., W. Dragowska, et al. (1994). "Evidence for a mitotic clock in human hematopoietic stem cells: loss of telomeric DNA with age." Proc Natl Acad Sci U S A 91(21): 9857-60. Vaziri, H., F. Schachter, et al. (1993). "Loss of telomeric DNA during aging of normal and trisomy 21 human lymphocytes." Am J Hum Genet 52(4): 661-7. Verheij, M., R. Bose, et al. (1996). "Requirement for ceramide-initiated SAPK/JNK signalling in stress-induced apoptosis." Nature 380(6569): 75-9. Vulliamy, T., A. Marrone, et al. (2002). "Association between aplastic anaemia and mutations in telomerase RNA." Lancet 359(9324): 2168-70. Vulliamy, T., A. Marrone, et al. (2001). "The RNA component of telomerase is mutated in autosomal dominant dyskeratosis congenita." Nature. 413(6854): 432-5. Vulliamy, T. J., A. Walne, et al. (2005). "Mutations in the reverse transcriptase component of telomerase (TERT) in patients with bone marrow failure." Blood Cells Mol Dis 34(3): 257-63. Wang, J., H. Feng, et al. (2005). "Human telomerase reverse transcriptase immortalizes bovine lens epithelial cells and suppresses differentiation through regulation of the ERK signaling pathway." J Biol Chem 280(24): 22776-87. Wang, J., L. Y. Xie, et al. (1998). "Myc activates telomerase." Genes Dev. 12(12): 176974. Wang, J. C., J. K. Warner, et al. (2005). "Dissociation of telomerase activity and telomere length maintenance in primitive human hematopoietic cells." Proc Natl Acad Sci U S A 102(40): 14398-403. Wang, S., C. Hu, et al. (2007). "Transcriptional silencing of a novel hTERT reporter locus during in vitro differentiation of mouse embryonic stem cells." Mol Biol Cell 18(2): 66977. Wang, S. and J. Zhu (2003). "Evidence for a relief of repression mechanism for activation of the human telomerase reverse transcriptase promoter." J Biol Chem. 278(21): 18842-50. Epub 2003 Feb 28. Wang, X., S. C. Wong, et al. (1998). "Evidence of cisplatin-induced senescent-like growth arrest in nasopharyngeal carcinoma cells." Cancer Res 58(22): 5019-22. Wang, Z., S. Kyo, et al. (2002). "Tamoxifen regulates human telomerase reverse transcriptase (hTERT) gene expression differently in breast and endometrial cancer cells." Oncogene. 21(22): 3517-24. Wang, Z., S. Kyo, et al. (2000). "Progesterone regulates human telomerase reverse transcriptase gene expression via activation of mitogen-activated protein kinase signaling pathway." Cancer Res. 60(19): 5376-81.

94

Watson, J. D. (1972). "Origin of concatemeric T7 DNA." Nat New Biol. 239(94): 197-201. Weng, N. P., B. L. Levine, et al. (1996). "Regulated expression of telomerase activity in human T lymphocyte development and activation." J Exp Med 183(6): 2471-9. Wenz, C., B. Enenkel, et al. (2001). "Human telomerase contains two cooperating telomerase RNA molecules." Embo J. 20(13): 3526-34. Wetterau, L. A., M. J. Francis, et al. (2003). "Insulin-like growth factor I stimulates telomerase activity in prostate cancer cells." J Clin Endocrinol Metab. 88(7): 3354-9. Wick, M., D. Zubov, et al. (1999). "Genomic organization and promoter characterization of the gene encoding the human telomerase reverse transcriptase (hTERT)." Gene. 232(1): 97-106. Wooten, L. G. and B. Ogretmen (2005). "Sp1/Sp3-dependent regulation of human telomerase reverse transcriptase promoter activity by the bioactive sphingolipid ceramide." J Biol Chem. 280(32): 28867-76. Epub 2005 Jun 10. Wu, G., W. H. Lee, et al. (2000). "NBS1 and TRF1 colocalize at promyelocytic leukemia bodies during late S/G2 phases in immortalized telomerase-negative cells. Implication of NBS1 in alternative lengthening of telomeres." J Biol Chem 275(39): 30618-22. Wu, K. J., C. Grandori, et al. (1999). "Direct activation of TERT transcription by c-MYC." Nat Genet. 21(2): 220-4. Wu, L. D., Y. Z. Chen, et al. (2002). "[Study on telomerase activity and expression of hTERT, c-myc and bcl-2 during terminal differentiation of HL-60 cells induced by retinoic acid]." Zhongguo Shi Yan Xue Ye Xue Za Zhi 10(5): 395-9. Wu, P., L. Meng, et al. (2005). "Role of hTERT in apoptosis of cervical cancer induced by histone deacetylase inhibitor." Biochem Biophys Res Commun 335(1): 36-44. Wu, Y. Y., A. M. Hruszkewycz, et al. (2000). "Limitations on the quantitative determination of telomerase activity by the electrophoretic and ELISA based TRAP assays." Clin Chim Acta 293(1-2): 199-212. Xu, D., S. Erickson, et al. (2000). "Interferon alpha down-regulates telomerase reverse transcriptase and telomerase activity in human malignant and nonmalignant hematopoietic cells." Blood. 96(13): 4313-8. Xu, D., A. Gruber, et al. (1999). "Suppression of telomerase reverse transcriptase (hTERT) expression in differentiated HL-60 cells: regulatory mechanisms." Br J Cancer 80(8): 1156-61. Xu, D., N. Popov, et al. (2001). "Switch from Myc/Max to Mad1/Max binding and decrease in histone acetylation at the telomerase reverse transcriptase promoter during differentiation of HL60 cells." Proc Natl Acad Sci U S A 98(7): 3826-31.

95

Yamaguchi, H. (2007). "Mutations of telomerase complex genes linked to bone marrow failures." J Nippon Med Sch 74(3): 202-9. Yan, P., J. M. Coindre, et al. (1999). "Telomerase activity and human telomerase reverse transcriptase mRNA expression in soft tissue tumors: correlation with grade, histology, and proliferative activity." Cancer Res 59(13): 3166-70. Yang, H., S. Kyo, et al. (2001). "Autocrine transforming growth factor beta suppresses telomerase activity and transcription of human telomerase reverse transcriptase in human cancer cells." Cell Growth Differ. 12(2): 119-27. Yi, X., J. W. Shay, et al. (2001). "Quantitation of telomerase components and hTERT mRNA splicing patterns in immortal human cells." Nucleic Acids Res 29(23): 4818-25. Yi, X., V. M. Tesmer, et al. (1999). "Both transcriptional and posttranscriptional mechanisms regulate human telomerase template RNA levels." Mol Cell Biol. 19(6): 398997. Yin, L., A. K. Hubbard, et al. (2000). "NF-kappa B regulates transcription of the mouse telomerase catalytic subunit." J Biol Chem. 275(47): 36671-5. Yu, C. C., S. C. Lo, et al. (2001). "Telomerase is regulated by protein kinase C-zeta in human nasopharyngeal cancer cells." Biochem J. 355(Pt 2): 459-64. Yu, G. L. and E. H. Blackburn (1991). "Developmentally programmed healing of chromosomes by telomerase in Tetrahymena." Cell. 67(4): 823-32. Zermati, Y., S. Fichelson, et al. (2000). "Transforming growth factor inhibits erythropoiesis by blocking proliferation and accelerating differentiation of erythroid progenitors." Exp Hematol 28(8): 885-94. Zhang, P., S. L. Chan, et al. (2003). "TERT suppresses apoptotis at a premitochondrial step by a mechanism requiring reverse transcriptase activity and 14-3-3 protein-binding ability." Faseb J 17(6): 767-9. Zhang, W., M. A. Piatyszek, et al. (1996). "Telomerase activity in human acute myelogenous leukemia: inhibition of telomerase activity by differentiation-inducing agents." Clin Cancer Res 2(5): 799-803. Zhou, X. Z. and K. P. Lu (2001). "The Pin2/TRF1-interacting protein PinX1 is a potent telomerase inhibitor." Cell 107(3): 347-59. Zhu, X., R. Kumar, et al. (1996). "Cell cycle-dependent modulation of telomerase activity in tumor cells." Proc Natl Acad Sci U S A 93(12): 6091-5. Zijlmans, J. M., U. M. Martens, et al. (1997). "Telomeres in the mouse have large interchromosomal variations in the number of T2AG3 repeats." Proc Natl Acad Sci U S A 94(14): 7423-8.

96

Zimmermann, S., S. Glaser, et al. (2004). "Effects of telomerase modulation in human hematopoietic progenitor cells." Stem Cells 22(5): 741-9. Zou, L., P. H. Zhang, et al. (2005). "Transcript regulation of human telomerase reverse transcriptase by c-myc and mad1." Acta Biochim Biophys Sin (Shanghai). 37(1): 32-8.

97

98

ANNEXES AUTRES PUBLICATIONS

99

Research Article

Cell Adhesion Regulates CDC25A Expression and Proliferation in Acute Myeloid Leukemia
Anne Fernandez-Vidal, Loc Ysebaert, Christine Didier, Remy Betous, Fabienne De Toni, 2 2 4 4 Nas Prade-Houdellier, Cecile Demur, Marie-Odile Contour-Galcera, Gregoire P. Prevost, 3 1 1 1 Bernard Ducommun, Bernard Payrastre, Claire Racaud-Sultan, and Stephane Manenti
1 Centre de Physiopathologie Toulouse-Purpan, Institut National de la Sante et de la Recherche Medicale U563-IFR30, Departement Oncogenese et Signalisation dans les cellules hematopoetiques, and 2Service dHematologie Biologique, CHU Purpan; ` 3 Laboratoire de Biologie cellulaire et Moleculaire du Controle de la Proliferation, Centre National de la Recherche Scientifique UMR 5088-IFR109, Universite Paul Sabatier, Toulouse, France; and 4IPSEN, Institut Henri Beaufour, Les Ulis, France

Abstract
The effects of cell adhesion on leukemia cell proliferation remain poorly documented and somehow controversial. In this work, we investigated the effect of adhesion to fibronectin on the proliferation of acute myeloid leukemia (AML) cell lines (U937 and KG1a) and CD34+ normal or leukemic primary cells. We observed an increased rate of proliferation of AML cells when adhered to fibronectin, concomitant with accelerated S-phase entry and accumulation of CDC25A. Conversely, normal CD34+ cell proliferation was decreased by adhesion to fibronectin with a concomitant drop in CDC25A expression. Importantly, we showed that both small interfering RNA (siRNA)mediated CDC25A down-regulation and a recently developed CDC25 pharmacologic inhibitor impaired this adhesion-dependent proliferation, establishing a functional link between CDC25A accumulation and adhesion-dependent proliferation in leukemic cells. CDC25A accumulation was found only slightly dependent on transcriptional regulation and essentially due to modifications of the proteasomal degradation of the protein as shown using proteasome inhibitors and reverse transcription-PCR. Interestingly, CDC25A regulation was Chk1 dependent in these cells as suggested by siRNA-mediated down-regulation of this protein. Finally, we identified activation of the phosphatidylinositol 3-kinase/ Akt pathway as an adhesion-dependent regulation mechanism of CDC25A protein expression. Altogether, our data show that in leukemic cells adhesion to fibronectin increases CDC25A expression through proteasome- and Chk1-dependent mechanisms, resulting in enhanced proliferation. They also suggest that these adhesion-dependent proliferation properties of hematopoietic cells may be modified during leukemogenesis. (Cancer Res 2006; 66(14): 7128-35)

Introduction
Eukaryotic cells need interactions with their microenvironment to regulate their proliferation. Adhesion to the extracellular matrix (ECM) has been extensively described as a key positive regulator of cell proliferation in various cell types (1), although in some cases

Requests for reprints: Stephane Manenti, Centre de Physiopathologie Toulouse Purpan, Institut National de la Sante et de la Recherche Medicale U563-IFR30, Departement Oncogenese et Signalisation dans les cellules hematopoetiques, CHU ` Purpan, BP3028, 31024 Toulouse Cedex 3, France. Phone: 33-562-74-45-24; Fax: 33-56274-45-58; E-mail: stephane.manenti@toulouse.inserm.fr. I2006 American Association for Cancer Research. doi:10.1158/0008-5472.CAN-05-2552

interaction with the ECM may inhibit cell proliferation (2, 3). Various cell signaling pathways have been involved in this regulation (4). In response to these signal transduction pathways, cyclin D1, p21Cip1, and p27Kip1 appear as the main G1 cell cycle regulators downstream of cell adhesion and can be considered as molecular sensors of extracellular stimuli (5, 6). Both transcriptional and post-translational regulations of these proteins have been described in this context. The microenvironment of hematopoietic stem cells and progenitors plays a crucial role in the normal and leukemic hematopoiesis in the bone marrow. Stimulation by various cytokines and direct interaction with the ECM and with neighbor cells determine differentiation and proliferation cues at the progenitor cells levels (7). For instance, a4 integrins regulate the proliferation/differentiation balance of hematopoietic progenitors (8). However, the functional and molecular effect of integrin-dependent cell adhesion to ECM on proliferation has not been extensively investigated in these cells. In normal human immature CD34+ cells, adhesion to fibronectin, a major component of the ECM in the bone marrow, inhibits cell proliferation (9, 10). Up-regulation of the cyclindependent kinase (CDK) inhibitor p27Kip1 was correlated with this inhibition. By contrast, this is apparently not the case in more mature (CD34) cells adhered to fibronectin (11). In acute myeloid leukemia (AML) cells, the role of the adhesion to the ECM on cell proliferation remains unclear, although in one study the adhesion to fibronectin of AML cells from some patients was found to act in synergy with stem cell factor (SCF) to induce proliferation (12). CDC25A is a dual-specificity phosphatase mostly involved in the G1-S transition of the cell cycle, although its implication in mitosis has also been recently described (13). Dephosphorylation of Thr14 and Tyr15 residues on CDK2 by CDC25A leads to the activation of the CDK2/cyclin E complex and participates to the G1-S-phase transition. To this respect, CDC25A can be considered as a key regulator of the G1 phase and as a potential sensor for extracellular stimuli. Still, major regulatory factors of this protein in response to mitogenic extracellular stimuli have not been extensively documented. E2F-dependent transcriptional regulation of CDC25A was reported and found necessary for efficient E2F-dependent S-phase entry (14). In an other study, this transcriptional regulation by E2F was described as a serum-dependent mechanism (15). Transcriptional regulation of CDC25A by c-myc was also reported, although the physiologic context of this process remains unclear (16, 17). In addition to transcriptional mechanisms, the CDC25A protein can be subjected to proteolytic processing by the proteasome. Double-strand breaks induced on DNA by genotoxic compounds or by radiations lead to the checkpoint kinase Chk1 activation and

Cancer Res 2006; 66: (14). July 15, 2006

7128

www.aacrjournals.org

Cell AdhesionDependent Regulation of CDC25A Differentiation of CD34+ normal progenitors in vitro. CD34+ cells were cultured and differentiated in 12-well plates in IMDM plus 10% FCS supplemented with SCF (100 ng/mL), IL-3 (5 ng/mL), FL (100 ng/mL), and GM-CSF (10 ng/mL) for myelomonocytic differentiation. Cells were initially seeded at 400,000/mL and diluted with fresh medium and cytokines every 2 days. Expression of the differentiation markers (CD34, CD38, CD33, and CD11b) was assessed by flow cytometry (200,000 cells used). For each time point, 300,000 cells were used for Western blot analysis. Transfection of small interfering RNA. Transfection experiments were done using the Amaxa nucleofection technology protocols (Amaxa, Cologne, Germany). U937 cells were grown until subconfluence (1 106/mL) and 6 106 cells were resuspended in 100 AL Amaxa solution V. Eight picomoles of Chk1 small interfering RNA (siRNA; Qiagen, Valencia, CA), CDC25A SMARTpool, Akt1 siRNA, or scramble siRNA (Dharmacon, Lafayette, CO) were added and cells were immediately transfected using the Amaxa nucleofector device (program V-01 for U937 and C-16 for Jurkat cells). For Akt, a second transfection was done in the same conditions 24 hours after the first one to improve the siRNA efficiency. Cells were subsequently resuspended in normal medium culture at 106/mL and processed after 8 (U937) or 24 ( Jurkat) hours for relevant experiments. Adhesion experiments were usually done at 0.5 106 cells/mL in the presence of 1% FCS. Western blot analysis and immunoprecipitation. Cells (2 106) were usually reduced in 75 AL Laemmli sample buffer, sonicated for 6 seconds, and boiled for 3 minutes. Proteins were then resolved on SDS-PAGE and transferred to nitrocellulose membrane (Hybond-C Super, Amersham Pharmacia Biotech, Piscataway, NJ). Saturation of the membrane was done for 1 hour in TBS with Tween 0.05% (TBS-T) containing 5% nonfat milk. Membranes were blotted with proper antibodies overnight at 4jC or for 1 hour at room temperature, washed thrice with TBS-T, and incubated for 30 minutes with horseradish peroxidasecoupled secondary antibody (Cell Signaling). After three additional washes, detection was achieved with Pierce Supersignal chemiluminescent substrate (Pierce, Rockford, IL). Antibodies were used at the following dilutions: anti-CDC25A (1:200), anticyclin A (1:1,000), anti-Chk1 (1:1,000), anti-Akt1/Akt2 (1:1,000), antiCDK2 (1:500), anti-phospho-Ser473-Akt (1:1,000), anti-phospho-Cdc2 (1:500), anti-phospho-p38 (1:500), and anti-actin (1:10,000). Western blot quantifications were done with the Gene Tools software from SynGene (Cambridge, United Kingdom). For CDK2 immunoprecipitation experiments, 5 106 cells were lysed in 1 mL immunoprecipitation buffer [50 mmol/L Tris (pH 8), 150 mmol/L NaCl, 3 mmol/L EGTA, 1% NP40, 50 mmol/L NaF, 1 mmol/L sodium orthovanadate, 10 mmol/L h-glycerophosphate, protease inhibitors] and kept for 20 minutes on ice. After centrifugation at 14,000 rpm for 10 minutes at 4jC, the supernatant was incubated with 5 Ag of the anti-CDK2 antibody and 40 AL protein A-Sepharose overnight at 4jC. Following two washes with the immunoprecipitation buffer, the immunoprecipitated material was resuspended in 60 AL Laemmli sample buffer, boiled for 3 minutes, and analyzed by Western blot with the anti-Tyr15 phosphorylated CDC2 antibody (1:1,000), which also recognizes the Tyr15 phosphorylated form of CDK2. Real-time PCR. Total RNA was extracted by guanidine isothiocyanatephenol-chloroform extraction ready-to-use TRIzol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer. Extraction was done from 1 107 cells in 1 mL TRIzol solution. cDNA was generated from 5 Ag RNA, with the SuperScript reverse transcriptase from the SuperScript III FirstStrand Synthesis System for Reverse Transcription-PCR (RT-PCR; Invitrogen) following the manufacturers suggestions. The PCR mixture containing 5 AL cDNA reactant and 0.5 Amol/L of each primer in a final volume of 25 AL prepared using the quantitative PCR Mastermix Plus SYBR Green I (Eurogentec, Seraing, Belgium). In negative controls, the reverse transcriptase was replaced by H2O. The primer sequences were as follows: CDC25A 5-ACCGTCACTATGGACCAGC-3 (sense; 0.5 Amol/L) and 5-TTCAGAGCTGGACTACATCC-3 (antisense; 0.5 Amol/L) and glyceraldehyde-3phosphate dehydrogenase (GAPDH) 5-TGCACCACCAACTGCTTAGC-3 (sense) and 5-GGCATGGACTGTGGTCATGAG-3 (antisense; 0.3 Amol/L). To confirm product purity, melting curves were checked and in some cases complemented by 1% agarose gel electrophoresis.

subsequent CDC25A phosphorylation, triggering its degradation by the proteasome through a SCFh-TrCP-dependent mechanism ( for a review, see ref. 18). Although this process was essentially described in response to genotoxic stress, recent studies suggest that this ubiquitin-proteasome system could also be involved in stress-independent CDC25A regulation during normal cell cycle (19). The question of how CDC25A levels are regulated seems of outstanding interest in cancer because overexpression of the protein has been described in various cancer cell types. In these studies, post-translational mechanisms of regulation have been found involved in the stabilization of the protein (20). In this work, we investigated how cell adhesion to fibronectin influences the proliferation of AML cells, and we established the importance of CDC25A regulation in this context. In these malignant cells, adhesion to fibronectin up-regulates CDC25A expression through regulation of its degradation, leading to increased proliferation.

Materials and Methods

Cytokines, antibodies, and pharmacologic inhibitors. The phosphatidylinositol 3-kinase (PI3K) inhibitors LY294002 and wortmannin and the mammalian target of rapamycin (mTOR) inhibitor rapamycin were purchased from Sigma (St. Louis, MO). The p38 mitogen-activated protein kinase (MAPK) inhibitor SB203580 and the proteasome inhibitors MG132 and lactacystin were purchased from Calbiochem (La Jolla, CA). UCN-01 was provided by the National Cancer Institute (Rockville, MD). SCF, interleukin (IL)-3, Flt3 ligand (FL), and granulocyte-macrophage colony-stimulating factor (GM-CSF) were from R&D Systems (Minneapolis, MN). Antibodies used were monoclonal anti-CDC25A, monoclonal anticyclin A, monoclonal anti-Chk1, polyclonal anti-CDK2, and polyclonal anti-Akt1/Akt2 (Santa Cruz Biotechnology, Santa Cruz, CA). Polyclonal anti-phospho-Ser473-Akt, anti-phospho-p38, and anti-phospho-CDC2 were from Cell Signaling (Beverly, MA). Actin monoclonal antibody (1:10,000) was from Sigma. Cell lines and culture conditions. Leukemic immature KG1a and more mature U937 and HL-60 cell lines were purchased from the German Collection of Microorganisms and Cell Cultures (Braunschweig, Germany) and grown in Iscoves modified Dulbeccos medium (IMDM) plus 20% FCS and RPMI plus 10% FCS, respectively, in the presence of 100 units/mL penicillin/ 100 Ag/mL streptomycin at 37jC and 5% CO2. Jurkat cells were purchased from the German Collection of Microorganisms and Cell Cultures and grown in RPMI plus 10% FCS. Normal bone marrow CD34+ hematopoietic cells were harvested from healthy volunteers upon informed consent after positive selection of CD34-expressing cells by the means of immunomagnetic separation column (Miltenyi Biotech, Bergisch Gladbach, Germany). The purity of CD34+ cells reached 85% to 98% according to flow cytometry analysis. Leukemic blasts from patients with newly diagnosed AML and referred to the Toulouse Purpan University Hospital were obtained upon informed consent from bone marrow aspirates by Ficoll-Hypaque density gradient centrifugation. Frozen cells were thawed in EGM2 medium (ref. 21; Cambrex, Walkersville, MD) supplemented with SCF (100 ng/mL), IL-3 (5 ng/mL), FL (100 ng/mL), and GM-CSF (10 ng/mL). Cells were left at 1 106/mL for 24 hours in these conditions during which moderate cell death occurred (10-40% of the cells depending of the patient). At that time, cells were processed for the proliferation experiments described in Results. Apoptotic cells were detected with the Annexin V-FITC detection kit from BD PharMingen (San Diego, CA) used as recommended by the manufacturer. Fibronectin matrix preparation and 3-(4,5-dimethylthiazol-2-yl)2,5-diphenyltetrazolium bromide proliferation assay. Six-well culture dishes were coated overnight at 4jC with 40 Ag/mL human fibronectin (Roche Molecular Biochemicals, Mannheim, Germany) in a final volume of 1 mL in PBS. Each well was rinsed twice with PBS before seeding the cells at relevant concentration (usually 0.5 106/mL for proliferation conditions). Quantification of living cells was quantified by 3-(4,5-dimethylthiazol-2-yl)2,5-diphenyltetrazolium bromide (MTT) assay (Sigma).

www.aacrjournals.org

7129

Cancer Res 2006; 66: (14). July 15, 2006

Cancer Research Bromodeoxyuridine incorporation assay. The S phase of the cell cycle was assayed by measuring the incorporation of bromodeoxyuridine (BrdUrd). Cells were pulse labeled with 10 Amol/L BrdUrd for 30 minutes, washed with PBS, and fixed in cold 70% ethanol for 20 minutes. BrdUrd detection was done by use of the BrdUrd staining kit from BD PharMingen (FITC-conjugated BrdUrd antibody set). The cells were then analyzed by flow cytometry with a FACScan cytometer (Becton Dickinson, San Diego, CA), and the percentage of cells synthesizing DNA was determined.

Results
Adhesion to fibronectin increases AML cell proliferation. To establish the effect of integrin-dependent adhesion on AML cell proliferation, U937 and KG1a cell lines were grown in suboptimal (1% FCS) conditions and maintained in suspension or adhered on a fibronectin matrix. Cells were counted daily by using a MTT colorimetric method. The results of these experiments show that after 3 days cell adhesion to fibronectin significantly increased the

Figure 1. Adhesion to fibronectin increases the proliferation rate of AML cells. A, U937 and KG1a cell lines were grown in the presence of 1% and 10% FCS, respectively, in normal suspension conditions (Sp; gray columns ) or in the presence of a fibronectin matrix (Adh ; white columns ). Viable cells were counted daily by the MTT method. The day 3/day 0 ratio was calculated from the corresponding absorbance values. Columns, mean of four independent experiments done in triplicate; bars, SD. **, P < 0.01; ***, P < 0.001, compared with suspension conditions. U937 and KG1a cells were left in suspension or adherent for 14 hours, and cyclin A expression was analyzed by Western blot after that time. Densitometric quantification of the Western blot signal was done, and the ratio between the CDC25A and the corresponding actin value was indicated for each fraction. B, AML blasts from patients were processed as described in Materials and Methods and used for proliferation experiment. The number of cells was counted each 48 hours, and the proliferation rate was calculated as in (A) and compared in the suspension (gray columns ) and adhesion (white columns ) culture conditions. Except for patient 1, in which cells underwent proliferation at day 1, proliferation between days 3 and 5 is shown for the other three patients.

proliferation rate of both U937 and KG1a leukemic cells (Fig. 1A). Annexin V-FITC apoptosis detection experiments were done as a control, indicating no significant difference of apoptosis between adherent and nonadherent cells (data not shown). BrdUrd incorporation experiments indicated an increase of 21.4 F 2% (two experiments) and 16.5 F 1.8% (three experiments) in S-phase entry of U937 and KG1a cells, respectively, on adhesion to fibronectin. To confirm this higher proliferation rate at the molecular level, we checked the expression of cyclin A, which is considered as a biochemical marker of the S-phase entry during the cell cycle. As can be seen in Fig. 1A (insets), cyclin A expression was significantly increased (1.9- and 2.7-fold in U937 and KG1a, respectively) when cells were adhered to fibronectin for 14 hours, consistent with the positive effect of cell adhesion on proliferation. To confirm these data, we did similar experiments with AML blasts from patients. Four patients with an immature phenotype (CD34+, M1) were used for this experiment (see results in Fig. 2 for justification). Cells were thawed in EGM2 medium in the presence of SCF, GM-CSF, IL-3, and FL, placed on a fibronectin matrix or left in suspension, and counted every 48 hours. As can be seen in Fig. 1B, AML blasts from three of these patients showed increased proliferation when adhered to fibronectin, confirming the data obtained with the established cell lines. Altogether, these data show that cell adhesion to fibronectin increases the proliferation of AML cells. Adhesion to fibronectin differently affects the proliferation of normal hematopoietic cells depending on their maturation state. To establish whether the above-mentioned adhesiondependent proliferation increase was specific of AML cells, we checked the effect of cell adhesion on the proliferation of normal hematopoietic progenitors from healthy donors. For this purpose, CD34+ normal hematopoietic cells were grown for 2 weeks in cytokine-supplemented medium conditions, allowing their differentiation along with the myelomonocytic pathway. This maturation process was conducted either in suspension or in adhesion conditions on fibronectin. The proliferation rate and the differentiation state were monitored every 2 days by direct counting of the living cells and by fluorescence-activated cell sorting (FACS) analysis, respectively. The results of these experiments indicate that proliferation of immature CD34+ cells is inhibited on binding to the fibronectin matrix (Fig. 2, top). By contrast, adhesion to fibronectin of more mature CD34 cells stimulated their proliferation (Fig. 2, bottom). These data suggest that the effect of integrin-dependent adhesion on cell proliferation depends on the differentiation stage of normal hematopoietic cells. They also point to an opposite behavior of normal and leukemic (see results with KG1a cell line and blasts from patients in Fig. 1) immature CD34+ cells in terms of adhesion-dependent cell proliferation. Adhesion to fibronectin up-regulates CDC25A protein level in AML cells. Cyclin D1, p27Kip1, and p21Cip1 are key cell cycle regulators involved in extracellular signals integration during the G1 phase of the cell cycle. To investigate the mechanisms of this anchorage-dependent proliferation increase, we checked the expression of these proteins by Western blot. In the experimental conditions used for proliferation studies, we did not observe any significant modification of the cellular expression levels for these proteins (data not shown). By contrast, we observed a dramatic increase of the dual-specificity phosphatase CDC25A, an other important player of the G1-S transition (Fig. 3A). This process was observed in the two previously described AML cell lines U937 and KG1a but also in the myeloblastic CD34 HL-60 cell line. These

Cancer Res 2006; 66: (14). July 15, 2006

7130

www.aacrjournals.org

Cell AdhesionDependent Regulation of CDC25A

adhesion on CDC25A protein level is clearly different from one cell type to the other. On adhesion to fibronectin, CDC25A increased in Jurkat and RL cells, decreased in Karpas 422 cells, and remained low and apparently unchanged in Karpas and ES cells. These data indicate that adhesion-dependent regulation of CDC25A is not restricted to AML cells and they confirm that this regulation may lead to opposite effects on CDC25A level. CDC25A accumulation is involved in adhesion-dependent proliferation. We then addressed the question of the functional relationship between adhesion-dependent CDC25A expression and increased cell proliferation. To this purpose, we down-regulated CDC25A expression in U937 cells by a siRNA strategy and checked the adhesion-dependent cyclin A expression by Western blot. As observed in Fig. 4A, a moderate siRNA-induced down-regulation of CDC25A significantly decreased the cyclin A protein level in adherent cells. These data suggest that modifications of CDC25A

Figure 2. Adhesion to fibronectin inhibits or stimulates the proliferation of normal hematopoietic cells depending on their maturation state. Adherent (white columns ) or suspension (gray columns ) normal hematopoietic CD34+ cells were differentiated in vitro as described in Materials and Methods. The proliferation was compared at two different stages (CD34+ and CD34) of the differentiation process. Results from two independent donors (donors 1 and 2). The cell number 3 days after seeding was compared for the adhesion and suspension conditions. Results are expressed in cell number/mL. For donor 1, the proliferation experiments were done between days 3 and 5 and days 10 and 12 for CD34+ and CD34 cells, respectively. For donor 2, proliferations were done between days 1 and 3 and days 8 and 10, respectively. This corresponded to similar stages of differentiation for the two donors as judged by FACS analysis of cell surface markers (data not shown).

results suggest for the first time the existence of an adhesiondependent regulation of CDC25A expression. We then tested the effect of cell adhesion on CDC25A expression in blasts from patients with immature CD34+ and M1 phenotype. As can be seen in Fig. 3B, blasts from patients 2 and 3 showed increased CDC25A expression when adhered to fibronectin, in good correlation with the proliferation experiments (see Fig. 1). Blasts from a third patient (patient 5), not tested for proliferation, also showed CDC25A expression increase in adhesive conditions. CDC25A level was found unchanged in patients 1 and 4 (data not shown). These data obtained with blasts from patients confirm that CDC25A protein is up-regulated in AML cells following adhesion to fibronectin. We then examined this regulation process in normal hematopoietic cells. As shown in Fig. 3C, CDC25A expression was high in CD34+ immature cells and dramatically reduced on adhesion to fibronectin. By contrast, the expression level of this protein was lower in more differentiated CD34 cells and significantly increased by the adhesion to fibronectin. These data suggest that (a) CDC25A protein level is regulated along the hematopoietic differentiation process and that cell interaction with the ECM may differently affect this expression and (b) up-regulation of CDC25A on adhesion is not maturation restricted in AML as opposed to their normal counterparts. We finally tested the adhesion-dependent regulation of CDC25A in other hematologic malignancies. Jurkat cells, the Karpas large cell anaplastic lymphoma cell line, the ES mantle lymphoma cell line, and the follicular lymphoma RL and Karpas 422 cell lines were tested to this respect. As shown in Fig. 3D, the effect of cell

Figure 3. CDC25A phosphatase expression is regulated by cell adhesion to fibronectin. A, AML cell lines were either adhered to fibronectin for 14 hours or left in suspension for the same time. CDC25A level was then visualized by Western blot. Membranes were then stripped and blotted again with an antibody against actin. B, blasts from patients were cultured on fibronectin or left in suspension for 14 hours and then processed for Western blotting analysis. In all cases, densitometric quantification of the Western blot signal was done, and the ratio between the CDC25A and the corresponding actin value was indicated for each fraction. C, normal hematopoietic CD34+ cells were subjected to a differentiation process in vitro as described before. At day 0 (CD34+ status) and day 7 (CD34 status), cells were collected and adhered to fibronectin matrix or left in suspension for 14 hours. CDC25A expression was then detected by Western blot analysis on the corresponding cellular extracts. These results were reproduced in at least three independent experiments for each cell line and twice for the primary cells. D, five cell lines, from various hematologic malignancies, were tested by Western blotting for CDC25A expression in suspension and adhesion conditions. Representative of two independent experiments. ALL, acute lymphoblastic leukemia; ML, mantle lymphoma; FL, follicular lymphoma; Ana.L, anaplastic lymphoma.

www.aacrjournals.org

7131

Cancer Res 2006; 66: (14). July 15, 2006

Cancer Research

Figure 4. CDC25A regulates adhesion-dependent proliferation of AML cells. A, U937 cells were transfected by scramble (control; scr) or CDC25A siRNA using the Amaxa nucleofection device. After 8 hours, cells were either adhered to fibronectin or left in suspension for 14 additional hours. CDC25A and cyclin A proteins levels were then detected by Western blot in the corresponding cellular extracts. Representative of three independent experiments. B, U937 cells were adhered or left in suspension for 14 hours in the presence (+) or absence () of the CDC25 inhibitor BN82685 (100 nmol/L). Western blot analysis of the cyclin A protein was then done. These experiments were done twice. In all cases, densitometric quantification of the Western blot signal was done, and the ratio between the CDC25A and the corresponding actin value was indicated for each fraction. C, cells were treated with BN82685 (300 nmol/L) for 4 hours and then processed for immunoprecipitation of CDK2 as described in Materials and Methods. The inactive form of CDK2 was then detected by Western blot with an anti-phospho-Tyr15 antibody, and total immunoprecipitated CDK2 was detected with the antibody used for the immunoprecipitation.

expression drive the augmentation of proliferation triggered by cell adhesion to fibronectin. To reinforce these observations, we made use of BN82685, a recently described CDC25 pharmacologic inhibitor (22). Because we address a question concerning the G1-S transition, we reasoned that in our experimental conditions the effect of this inhibitor should essentially concern the CDC25A phosphatase and not CDC25B or CDC25C, although these two isoforms are also inhibited in these conditions. When U937 cells were adhered to fibronectin in the presence of this inhibitor, we observed that cyclin A accumulation was impaired, in good correlation with the siRNA experiments (Fig. 4B). Altogether, these data show that CDC25A up-regulation is involved in the adhesion-dependent proliferation increase of AML cells. To confirm the effect of BN82685 on CDK2 activity, we did an immunoprecipitation of CDK2 from control and BN82685-treated cells and detected the inactive form of the kinase by Western blot with an antibody recognizing Tyr15-phosphorylated CDK2. As shown in Fig. 4C, BN82685 indeed increased the inhibitory phosphorylation state of CDK2 in these experiments. Cell typedependent variations of CDC25A mRNA on adhesion to fibronectin. To establish if transcriptional activation could control the CDC25A protein accumulation, we compared CDC25A mRNA levels in adherent and nonadherent AML cells. Quantitative RT-PCR was done for this purpose. GAPDH mRNA levels were used to normalize the measured values, and the results

were expressed in fold increase between adhesion and suspension conditions (Fig. 5). The results of these experiments indicate almost no variation of CDC25A mRNA in U937 and HL-60 cell lines (1.6 and 1.4, respectively) and a moderate increase in KG1a adherent cells (4.5). These data suggest that transcriptional increase and/or mRNA stabilization are probably not involved in CDC25A protein accumulation in U937 and HL-60 cell lines, although they possibly participate to this regulation in KG1a cells. In consequence, they also suggest some cell typedependent variations of this regulation. Involvement of Chk1 and proteasomal degradation in CDC25A turnover. We then investigated whether the previously described post-translational regulation of CDC25A was involved in the accumulation of this protein on adhesion. We focused our attention on proteasomal degradation of the protein and on the potential role of the checkpoint kinase Chk1 in this process. When AML cell lines were incubated with the proteasome inhibitor MG132 for 5 hours in growing culture conditions, an accumulation of CDC25A protein was observed in each cell line (Fig. 6A). When U937 cells were transfected with Chk1 siRNA, leading to major down-regulation of this protein, a significant increase of CDC25A expression was also observed (Fig. 6B). Altogether, these data suggest that phosphorylation by Chk1 and subsequent proteasomal degradation are key regulators of CDC25A protein in the absence of genotoxic stress in these cells. To confirm a role for Chk1 in the adhesion-dependent regulation of CDC25A, adherent U937 and KG1a cells were removed from the fibronectin matrix with EDTA in the presence of the Chk1 inhibitor UCN-01. First, removing the cells from the fibronectin matrix with EDTA induced rapid downregulation of the CDC25A protein, indicating that the adhesiondependent accumulation process was quickly reversible. As can be seen in Fig. 6C, the presence of UCN-01 impaired this CDC25A down-regulation, suggesting that Chk1 is actually involved in the adhesion-dependent regulation of the phosphatase. We then questioned the involvement of proteasome regulation in adhesion-dependent CDC25A increase. To this aim, the same experiment as in Fig. 6C was conducted in the presence of proteasome inhibitors instead of UCN-01. The results of these experiments are presented in Fig. 6D. The presence of proteasome inhibitors completely or partially inhibited this down-regulation depending on the cell type and on the pharmacologic inhibitor used. Altogether, these data suggest that Chk1 and the proteasome machinery are actually involved in the anchorage-dependent CDC25A regulation.

Figure 5. RT-PCR analysis of CDC25A expression in adhesion and suspension conditions. Total mRNA was prepared from adherent (14 hours) and nonadherent cells and quantitative RT-PCR was done to quantify CDC25A mRNA as described in Materials and Methods. The ratio for adherent versus suspension cells from the three cell lines was then deduced and plotted. Columns, mean of three independent experiments done in triplicate; bars, SD.

Cancer Res 2006; 66: (14). July 15, 2006

7132

www.aacrjournals.org

Cell AdhesionDependent Regulation of CDC25A

strongly impaired the adhesion-dependent CDC25A up-regulation, suggesting the participation of this pathway in this process. To further show the involvement of the PI3K pathway in CDC25A regulation, we aimed to perform siRNA-dependent Akt downregulation. Due to technical difficulties to use the U937 cells for this purpose, we used the Jurkat cell line found previously (see Fig. 3) to regulate CDC25A level through an adhesion-dependent mechanism. BrdUrd incorporation experiments done with this cell line showed increased DNA synthesis (17.7 F 2.5%) on cell adhesion to fibronectin. We then verified that the PI3K inhibitors LY294002 and wortmannin impaired the CDC25A protein level in this cell type (Fig. 8A). In good agreement, siRNA-mediated downregulation of the Akt protein led to reduced expression of CDC25A in adherent cells (Fig. 8B) despite incomplete (50%) downregulation of Akt protein in the conditions used. Altogether, these data suggest that interaction with fibronectin activates the PI3K/ Akt/mTOR pathway, which participates to CDC25A up-regulation in these conditions.

Discussion
In this work, we show that adhesion to fibronectin stimulates the proliferation of AML cell lines. This question had not been extensively addressed until now, although the proliferation state of leukemic cells in the bone marrow may be of outstanding
Figure 6. Proteasome- and Chk1-dependent regulation of CDC25A in AML cells. A, U937, HL-60, and KG1a cells were incubated with the proteasome inhibitor MG132 (MG ; 5 Amol/L) for 5 hours. CDC25A level was then detected by Western blot. B, U937 cells were transfected with scramble or Chk1 siRNA with the Amaxa nucleofection system. After 24 hours, cell extracts were prepared and Chk1 and CDC25A expression was detected by Western blot. C, U937 and KG1a cells were adhered to fibronectin for 14 hours, removed from the fibronectin matrix with EDTA (ReSp ), and left in suspension conditions in the presence of UCN-01 (UCN; 100 nmol/L) or with DMSO. D, as in (C ), U937, HL-60, and KG1a cells were adhered to fibronectin for 14 hours, removed from the fibronectin matrix with EDTA, and left in suspension conditions in the presence of the proteasome inhibitors MG132 (5 Amol/L) and lactacystin (Lac ; 5 Amol/L) or with DMSO (C ). Western blot analysis of the corresponding cellular fractions was then done to follow CDC25A expression. Representative data of at least two independent experiments done from each cell line. In all cases, densitometric quantification of the Western blot signal was done, and the ratio between the CDC25A and the corresponding actin value was indicated for each fraction.

PI3K/mTOR pathway is involved in adhesion-dependent CDC25A regulation. Because the p38 MAPK pathway was recently involved in CDC25A protein expression in T lymphocytes, we investigated p38 involvement in U937 and KG1a cell proliferation in suspension and adhesion conditions. The p38 inhibitor SB203580 did not affect CDC25A levels in any of the conditions or cell lines tested, suggesting that p38 MAPK is not a major regulator of CDC25A expression in AML (data not shown). A few reports from the literature suggest that the PI3K/mTOR pathway can regulate CDC25A level in certain conditions (23, 24). Therefore, we tested the effect of PI3K and mTOR inhibitors on anchorage-dependent CDC25A accumulation. First, we checked the effect of cell adhesion on the activity of the PI3K pathway. For this purpose, U937 were adhered to fibronectin for different times, and the phosphorylation state of protein kinase B (PKB)/Akt was detected by Western blot. As shown in Fig. 7A, rapid and sustained activation of this pathway was observed on adhesion of U937 cells to fibronectin. Interestingly, this adhesion-dependent activation of PKB/Akt was also observed in AML primary cells from patients (Fig. 7B). As shown in Fig. 7C and D, the PI3K inhibitors LY294002 and wortmannin and the mTOR-specific inhibitor rapamycin

Figure 7. PI3K pathway regulates adhesion-dependent CDC25A protein accumulation. A, U937 cells were adhered to fibronectin for different times or left in suspension, and CDC25A expression and Akt/PKB Ser473 phosphorylation state (P-Akt ) were probed by Western blot. B, blasts from AML patients were left in suspension or adhered to fibronectin for 14 hours, and Akt/PKB Ser473 phosphorylation and Akt protein level were then probed by Western blot analysis. C, U937 cells were adhered to fibronectin for 14 hours in the absence (C ) or presence of the PI3K inhibitors LY294002 at 10 (LY10 ) or 20 (LY20 ) Amol/L or wortmannin at 50 (W50 ) or 100 (W100 ) nmol/L. CDC25A expression and Akt/ PKB Ser473 phosphorylation state were then probed by Western blot. Representative of four independent experiments. D, U937 cells were left in suspension conditions or adhered to fibronectin in the absence or presence of the mTOR inhibitor rapamycin (Rap ; 10 nmol/L). Western blot analysis of CDC25A expression was then done. Representative of two independent experiments.

www.aacrjournals.org

7133

Cancer Res 2006; 66: (14). July 15, 2006

Cancer Research

Figure 8. PKB/Akt down-regulation affects CDC25A level in Jurkat cells. A, Jurkat cells were left in suspension or adhered to fibronectin for 14 hours in the absence (Cont ) or the presence of LY294002 (LY; 20 Amol/L) or wortmannin (Wort. ; 100 nmol/L). CDC25A expression, Akt protein level, and Akt/PKB Ser473 phosphorylation state were then probed by Western blot. Representative of three experiments. B, Jurkat cells were transfected with scramble (Scr ) or Akt siRNA as described in Materials and Methods and then adhered to fibronectin for 14 hours. Akt and CDC25A expression was then probed by Western blot analysis. Representative of two independent experiments.

importance for their response to therapeutic drugs and/or relapse of the disease. A synergistic effect of cell adhesion on SCF-mediated AML cell proliferation has been already reported in some cases of leukemic blasts from patients (12). Our experiments done on established cell lines and blasts from patients further reinforce these data and suggest that blast adhesion to fibronectin in the bone marrow may be a key variable amplifying the abnormal expansion of the leukemic clone. Of special interest are the results obtained with normal immature (CD34+) cells. Adhesion onto a fibronectin matrix inhibits their proliferation as already reported by others (9, 10). One of the important observations is that this phenotype apparently disappears along with the myeloid differentiation process and mirrors the kinetics of CD34 antigen expression. Indeed, a more mature phenotype corresponding to CD34 committed precursors showed accelerated proliferation on adhesion to fibronectin. Conversely, the maturation state, as assessed by French-American-British classification, does not seem anymore as a critical factor for the adhesion-dependent proliferation of leukemic cells, because both immature (CD34+) AML cells (KG1a cell line and blasts from patients) and U937, a more mature myelomonocytic (CD34) cell line, showed increased anchoragedependent proliferation. The opposite effects on proliferation triggered by adhesion observed between normal and leukemic CD34+ cells suggest that modifications of the proliferative response to cell adhesion could be a critical step occurring during the leukemogenic process. Experiments are presently being done in our laboratory to answer this question. An other important observation is that CDC25A expression was dramatically dependent on the adhesive status of the cells. To our knowledge, this is the first report of an integrin-dependent regulation of CDC25A. As a general matter, there are not many reports of extracellular signal-dependent variations of CDC25A protein level in the literature. Serum- and E2F-dependent increase

of CDC25A mRNA has been described, but the status of the protein was poorly documented in these studies (14, 15). Furthermore, our quantitative RT-PCR analysis of CDC25A mRNA does not argue for a major transcriptional regulation of the protein in our system. Our data are more in favor of post-translational regulation(s) of the protein by the proteasome as suggested by the experiments done with pharmacologic inhibitors of this machinery. However, the 4.5-fold increase of CDC25A mRNA in KG1a cells on adhesion may suggest that transcriptional or mRNA stability regulations are active in certain cell types. Further experiments are needed to confirm these hypotheses. One of the important questions ensuing from our data is the cell specificity of this adhesion-dependent CDC25A regulation. The experiments done with other hematologic malignancies suggest that this regulation is not specific of AML cells. Although we did not perform proliferation experiments with these different cell lines, one may notice that increased proliferation of Jurkat cells has been described in response to cell adhesion (see ref. 25), in good agreement with the CDC25A accumulation observed in these cells (see Fig. 3) and with the increased BrdUrd incorporation that we observed on cell adhesion. Finally, although preliminary experiments from our laboratory indicate that nonhematopoietic cell lines (HeLa and HEK293) do not significantly change their CDC25A expression on cell anchorage removal, we cannot rule out the existence of this regulation in such cell types. Our data also suggest the involvement of the PI3K pathway and of the Chk1 kinase in the adhesion-dependent regulation of CDC25A. We did not detect any adhesion-regulated modulation of Chk1 activity by Western blot with an anti-phospho-Ser345 antibody (data not shown). This does not rule out the existence of some other type(s) of regulation not detectable by this approach. Interestingly, recent publications describe the phosphorylation of Chk1 by Akt and the subsequent inactivation of the kinase by cytoplasmic sequestration following monoubiquitination (26, 27). This hypothesis would nicely fit with our observations. Indeed, cell anchorage triggers Akt activation in our system and PI3K inhibition impairs the adhesion-dependent CDC25A up-regulation. Interestingly, a few data from the literature already ascribed a role for the PTEN/PI3K (23) and/or mTOR (24) pathways in CDC25A protein accumulation. Furthermore, the PI3K/Akt/mTOR pathway is currently deregulated in AML cells from patients (28), and we recently described the mTOR kinase as an interesting therapeutic target in these pathologies (29). Whether an integrin/PI3K/Akt/ Chk1 pathway regulates CDC25A expression in AML cells is currently under study in our laboratory. As indicated in Results, a recent work (30) described a major role for CDC25A in the proliferation of lymphocytes. Theses authors described a direct phosphorylation of CDC25A by the p38 MAPK pathway in cytokine-starved lymphocytes and the subsequent down-regulation of the phosphatase. Although we observed some regulation of the p38 MAPK activity on cell adhesion in KG1a cells, this pathway is apparently not involved in the adhesion-dependent CDC25A regulation. This discrepancy argues for different mechanisms of CDC25A regulation downstream of various mitogenic signals (growth factors, cell adhesion, etc.) and/or a cell type dependent regulation of the phosphatase. Our data argue for the involvement of CDC25A in the adhesiondependent proliferation increase of AML cells. Whether cell adhesion to the ECM participates to the leukemogenic proliferation in vivo remains to be established, but if this was the case this would make of CDC25A an interesting therapeutic target for these

Cancer Res 2006; 66: (14). July 15, 2006

7134

www.aacrjournals.org

Cell AdhesionDependent Regulation of CDC25A

pathologies. Of outstanding interest is the fact that the expression of this protein seems regulated during the normal hematopoietic process and permanently under the positive (mature progenitors) or negative (immature progenitors) control of integrin functions.

Acknowledgments
Received 7/20/2005; revised 4/19/2006; accepted 5/19/2006.

Grant support: Institut National de la Sante et de la Recherche Medicale, Association Laurette Fugain grant R05027BB, and Association pour la Recherche sur le Cancer grant 3638. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. We thank Pierre Cavailles and Marie-Michele Duplantier for helpful expertise and discussions on real-time quantitative PCR, Martin Carroll for helpful discussions about the culture of blasts from patients, Christine Bezombes for providing us with lymphoblastic cell lines, and the Division of Cancer Treatment and Diagnosis, National Cancer Institute, for providing us with UCN-01.

References
1. Danen E, Yamada K. Fibronectin, integrins and growth control. J Cell Physiol 2001;189:113. 2. Meredith J, Takada Y, Fornaro M, Languino LR, Schwartz MA. Inhibition of cell cycle progression by the alternatively spliced integrin h1C. Science 1995;269:15702. 3. Fornaro M, Steger CA, Bennett AM, Wu JJ, Languino L. Differential role of h1C and h1A integrin cytoplasmic variants in modulating focal adhesion kinase, protein kinase B/AKT, and Ras/mitogen-activated protein kinase pathways. Mol Biol Cell 2000;11:223549. 4. Lee JW, Juliano R. Mitogenic signal transduction by integrin and growth factor receptor-mediated pathways. Mol Cells 2004;17:188202. 5. Assoian R. Anchorage-dependent cell cycle progression. J Cell Biol 1997;136:14. 6. Assoian RK, Schwartz MA. Coordinate signaling by integrins and receptor tyrosine kinases in the regulation of G1 phase cell-cycle progression. Genes Dev 2001;11:4853. 7. Verfaillie C. Adhesion receptors as regulator of hematopoietic process. Blood 1998;92:260912. 8. Arroyo AG, Yang JT, Rayburn H, Hynes RO. a4 Integrins regulate the proliferation/differentiation balance of multilineage hematopoietic progenitors in vivo . Immunity 1999;11:55566. 9. Hurley RW, McCarthy JB, Verfaillie CN. Direct adhesion to bone marrow stroma via fibronectin receptors inhibits hematopoietic progenitor proliferation. J Clin Invest 1995;96:5119. 10. Jiang Y, Prosper F, Verfaillie CM. Opposing effects of engagement of integrins and stimulation of cytokine receptors on cell cycle progression of normal human hematopoietic progenitors. Blood 2000;95:84654. 11. Dylla SJ, Deyle DM, Theunissen K, Padurean AM, Verfaillie CM. Integrin engagement-induced inhibition

of human myelopoiesis is mediated by proline-rich tyrosine kinase 2 gene products. Exp Hematol 2004;32: 36574. 12. Bendall LJ, Makrynikola V, Hutchinson A, Bianchi AC, Bradstock KF, Gottlieb DJ. Stem cell factor enhances the adhesion of AML cells to fibronectin and augments fibronectin-mediated anti-apoptotic and proliferative signals. Leukemia 1998;12:137582. 13. Mailand N, Podtelejnikov AV, Groth A, Mann M, Bartek J, Lukas J. Regulation of G2-M events by CDC25A through phosphorylation-dependent modulation of its stability. EMBO J 2002;21:591120. 14. Vigo E, Muller H, Prosperini E, et al. CDC25A phosphatase is a target of E2F and is required for efficient E2F-induced S phase. Mol Cell Biol 1999;19:637995. 15. Chen X, Prywes R. Serum-induced expression of the CDC25A gene by relief of E2F-mediated repression. Mol Cell Biol 1999;19:4695702. 16. Galaktionov K, Chen X, Beach D. CDC25A cell-cycle phosphatase as a target of c-myc. Nature 1996;382:5117. 17. Barre B, Vigneron A, Coqueret O. The STAT3 transcription factor is a target for the Myc and retinoblastoma proteins on the CDC25A promoter. J Biol Chem 2005;280:1567381. 18. Busino L, Chiesa M, Draetta GF, Donzelli M. CDC25A phosphatase: combinatorial phosphorylation, ubiquitylation and proteolysis. Oncogene 2004;23:20506. 19. Sorensen CS, Syljuasen RG, Bartek J, Lukas J. ATR, Claspin and the Rad9-1-Hus1 complex regulate Chk1 and CDC25A in the absence of DNA damage. Cell Cycle 2004;3:e359. 20. Loffler H, Syljuasen RG, Bartkova J, Worm J, Bartek J, Lukas J. Distinct mode of deregulation of the protooncogenic CDC25A phosphatase in human breast cancer cell lines. Oncogene 2003;22:806371.

21. Xu Q, Thompson JE, Carroll M. mTOR regulates cell survival after etoposide treatment in primary AML cells. Blood 2005;106:42618. 22. Brezak MC, Quaranta M, Contour-Galcera MO, et al. Inhibition of human tumor cell growth in vivo by an orally bioavailable inhibitor of CDC25 phosphatases. Mol Cancer Ther 2005;4:137887. 23. Seminario MC, Precht P, Wersto RP, et al. PTEN expression in PTEN-null leukaemic T cell lines leads to reduced proliferation via slowed cell cycle progression. Oncogene 2003;22:8195204. 24. Gao N, Flynn DN, Zhang Z, et al. G1 cell cycle progression and the expression of G1 cyclins are regulated by PI3K/Akt/mTOR/p70S6K1 signaling in human ovarian cancer cells. Am J Physiol Cell Physiol 2004;287:C28191. 25. Iwata S, Ohashi Y, Kamiguchi K, Morimoto C. h1Integrin-mediated cell signaling in T lymphocytes. J Dermatol Sci 2000;23:7586. 26. Puc J, Keniry M, Li HS, et al. Lack of PTEN sequesters Chk1 and initiates genetic instability. Cancer Cell 2005;7: 193204. 27. Puc J, Parsons R. PTEN loss inhibits CHK1 to cause double stranded-DNA breaks in cells. Cell Cycle 2005;4: e713. 28. Min YH, Eom JI, Cheong JW, et al. Constitutive phosphorylation of Akt/PKB protein in acute myeloid leukemia: its significance as a prognostic variable. Leukemia 2003;17:9957. 29. Recher C, Beyne-Rauzy O, Demur C, et al. Antileukemic activity of rapamycin in acute myeloid leukemia. Blood 2005;105:252734. 30. Khaled A, Bulavin D, Kittipatarin C, et al. Cytokinedriven cell cycling is mediated through CDC25A. J Cell Biol 2005;169:75563.

www.aacrjournals.org

7135

Cancer Res 2006; 66: (14). July 15, 2006

Leukemia (2006) 20, 12111216 & 2006 Nature Publishing Group All rights reserved 0887-6924/06 $30.00
www.nature.com/leu

ORIGINAL ARTICLE Expression of b-catenin by acute myeloid leukemia cells predicts enhanced clonogenic capacities and poor prognosis
L Ysebaert1,2, G Chicanne1, C Demur1,3, F De Toni1, N Prade-Houdellier1, J-B Ruidavets4, V Mansat-De Mas1,3, F Rigal-Huguet2, G Laurent1,2, B Payrastre1, S Manenti1 and C Racaud-Sultan1 ` INSERM U563, Centre de Physiopathologie Toulouse Purpan (CPTP), Departement Oncogenese et Signalisation dans les cellules Hematopoetiques, CHU Purpan, Toulouse Cedex, France; 2Service dHematologie clinique, CHU Purpan, Toulouse, France; 3 Service dHematologie biologique, CHU Purpan, Toulouse, France and 4INSERM U558, Faculte de Medecine de Toulouse, Toulouse, France
1

Activation of the Wnt/b-catenin pathway has recently been shown to be crucial to the establishment of leukemic stem cells in chronic myeloid leukemia. We sought to determine whether b-catenin was correlated to clonogenic capacity also in the acute myeloid leukemia (AML) setting. Eighty-two patients were retrospectively evaluated for b-catenin expression by Western blot. b-Catenin was expressed (although at various protein levels) in 61% of patients, and was undetectable in the remaining cases. In our cohort, b-catenin expression was correlated with the clonogenic proliferation of AML-colony forming cells (AML-CFC or CFU-L) in methylcellulose in the presence of 5637-conditioned medium, and more strikingly with self-renewing of leukemic cells, as assessed in vitro by a replating assay. In survival analyses, b-catenin appeared as a new independent prognostic factor predicting poor event-free survival and shortened overall survival (both with Po0.05). Furthermore, variations in b-catenin protein levels were dependent on post-transcriptional mechanisms involving the Wnt/ b-catenin pathway only in leukemic cells. Indeed, b-catenin negative leukemic cells were found to increase b-catenin in response to Wnt3a agonist in contrast to normal counterparts. Altogether, our data pave the way to the evaluation of Wnt pathway inhibition as a new rationale for eradicating the clonogenic pool of AML cells. Leukemia (2006) 20, 12111216. doi:10.1038/sj.leu.2404239; published online 11 May 2006 Keywords: Wnt; b-catenin; acute myeloid leukemia; prognosis; clonogenicity; CFU-L

different growth patterns in AML cells have emerged from some pioneer studies,6,7 but the ability to form colonies in methylcellulose allowed to gain more insight into the denition of true CFU-L. Later, these preliminary ndings were conrmed by classifying CFU-L based on autonomous growth or dependancy upon addition of exogenous growth factors.8 Wnt pathway controls self-renewal of hematopoietic stem cells (HSC) through regulation of b-catenin.911 In HSC, b-catenin also modulates survival and growth factor response both in vitro and in vivo, while inhibiting differentiation.1114 The importance of b-catenin in the maintenance of the HSC pool echoes in many hematological malignancies of lymphoid or myeloid origin.1521 As clonogenic test of AML cells at diagnosis are routinely performed in our lab for newly diagnosed patients, we tried to identify b-catenin expression in unsorted bulk cells to know whether it could correlate with enhanced clonogenic properties. We also studied the inuence of Wnt stimulation in the accumulation of b-catenin in normal CD34 or leukemic cells and along myelomonocytic maturation induced in vitro. We conrm that Wnt signaling is deregulated in AML as assessed by b-catenin expression (which is rapidly lost upon induction of differentiation in normal precursors), and is correlated to enhanced clonogenic proliferation of CFU-L. Moreover, b-catenin expression appears as a new prognostic factor independently correlated with relapse and shorter survival. Finally, the Wnt pathway could be regulated in AML, thus suggesting a different responsiveness to Wnt agonists of AML comparatively to normal counterparts.

Introduction
Wnt/b-catenin signaling is a morphogenic pathway that governs early embryo patterning and organogenesis in embryo development. Its implication in the self-renewal of normal adult stem cells raised major interest in the elds of regenerative medicine and cancer.13 As b-catenin, the main effector of the canonical Wnt pathway, was linked to clonogenic growth of various malignancies,45 we asked whether it could also inuence leukemic colony forming cells (CFU-L). These rare progenitors support self-renewal (in the minority) or limited differentiation (in the bulk of the disease) of acute myeloid leukemia (AML) blasts. The frequency of relapse in AML is thought to be due to the lack of eradication of the self-renewing part of AML cells, the so-called leukemic stem cells pool. In vitro assessment of
Correspondence: Dr L Ysebaert, Service dHematologie clinique, CHU Purpan, 1 place du Dr Baylac, 31059 Toulouse Cedex 9, France. E-mail: ysebaert.l@chu-toulouse.fr Received 3 September 2005; revised 17 March 2006; accepted 23 March 2006; published online 11 May 2006

Materials and methods

Primary cells and cell lines

Leukemic blasts from randomly selected 82 patients with newly diagnosed AML, were obtained upon informed consent from bone marrow aspirates by Ficoll-Hypaque density gradient centrifugation. This technique ensures high blastic cells percentages (median 92% of leukemic cells in the samples). Cells were either immediately processed or cryopreserved in Iscoves modied Dulbecco medium (IMDM) with 10% dimethylsulfoxyde and 50% fetal calf serum (FCS) for further handling. All patients received induction therapy (anthracycline days 15 and cytosine arabinoside continuously infused days 17), further consolidation therapy being given accordingly to French cooperative trials protocols, stratied by age. Normal bone marrow CD34 were sorted by the means of immunomagnetic separation column (Miltenyi Biotech, Germany). The purity of CD34 cells reached 8598% according to ow cytometry.

b-Catenin and clonogenic capacity in acute myeloid leukemia L Ysebaert et al

1212
Leukemic immature (KG1) and myelomonocytic (U937) cell lines were purchased from DSMZ (Germany), and grown in IMDM 10% FCS and RPMI 10% FCS, respectively, with penicillin/streptomycin at 371C and 5% CO2. Plating Efciency at 1 week) or without (PE1-) 10% human bladder carcinoma cell line 5637 conditioned medium (5637 CM). A total of 1 105 cells was then plated in 35 mm Petri dishes in triplicate (1 ml mixture/dish), and incubated at 371C and 5% CO2 humidied atmosphere. At day 7, the leukemic colonies (420 cells) and clusters (45 cells) were scored under an inverted microscope. As the growth of clusters or colonies depends on intrinsic blastic characters, clusters and colonies were referred as CFU-L. Cells were washed, counted and re-seeded at 1 105/ml in the same solid medium with 5637 CM for assessment of replating capacity (plating efciency at 2 weeks (PE2)) at day 15.

CD34 cells were maintained in liquid culture and differentiated in six-well plates in IMDM 10% FCS, supplemented with stem cell factor (SCF, 100 ng/ml), Interleukin 3 (IL-3, 10 ng/ ml), Flt3 ligand (FL, 50 ng/ml) and granulocyte macrophagecolony stimulating factor (GM-CSF, 50 ng/ml) for myelomonocytic differentiation (for 14 days) (all cytokines purchased from R&D Systems, Minneapolis, MN, USA). Input plating was 4 105/ml, and cells were counted and diluted with fresh medium with cytokines ( / Wnt ligands) every 2 days. When indicated, cells were cultured under the same conditions with recombinant Wnt3a (20 ng/ml) or Wnt5a (0.3 mg/ml) from R&D Systems. Choice of Wnt dosage was guided by manufacturers guidelines and by publication by Willert et al.,12 who used puried Wnt3a and demonstrated that accumulation of b-catenin is comparable between 20 and 1000 ng/ml. Expression of differentiation markers (CD34, CD38, CD33, CD11b and CD14) was assessed at day 3, 5, 7, 10 and 14 by ow cytometry. At each time point, 2 105 cells were also lysed for further blotting.

Differentiation of CD34 normal progenitors in vitro

Statistical analysis
Inuence of b-catenin expression was evaluated on patients clinical and biological characteristics using the w2 test and Fishers exact test for categorical variables. Survival curves were plotted using the product-limit method of KaplanMeier and compared with the log-rank test. MannWhitney test was carried out for comparison of medians from nonparametric data. Cox regression model was used for the identication of prognostic factors, those linked with RFS and OS with Po0.1 in univariate analysis were subsequently assessed in multivariate analysis (level for signicance: Po0.05)

Reverse transcriptase-polymerase chain reaction analysis

Total RNAs were extracted using Qiagen RNAeasy kit, according to manufacturers instructions, and were reverse-transcribed using the Superscript II Reverse Transcriptase (Invitrogen, Carlsbad, CA, USA). Of a 20 ml cDNA reaction 5 ml was used as template for amplication with the following specic primers. For human Wnt5a forward: 50 -CAAGGTGGGTGATGCCCT GAAGGAG-30 , and reverse 50 -CGTCTGCACGGTCTTGA ACTGGTCGTA-30 , for human b-catenin forward: 50 -AATCA GCTGGCCTGGTTTGA-3, and reverse 50 -GGCCAATCACAA TGCAAGTTC-30 , for human GAPDH forward: 50 -GGCTGA GAACGGGAAGCTTGTCAT-3, and reverse 5-CAGCCTTCTC CATGGTGGTGAAGA-30 .

Results

Expression of b-catenin in AML cells and leukemic cell lines

b-Catenin was found to be expressed in 61% of our 82 AML samples, as well as in nearly all tested leukemic cell lines (Figure 1), although at various expression levels. As expected with regards to its post-translational regulation by the Wnt pathway, b-catenin protein levels differed from mRNA levels as assessed by RT-PCR (Figure 1). Indeed, mRNA was constantly found by RT-PCR analysis, irrespective of protein levels, indicating that b-catenin is mainly regulated through translational and/or post-translational mechanisms in AML cells. Leukemic cells expressing b-catenin were uniformly classied as b-catenin for further correlation analyses, irrespective of the level of accumulation of the protein. On the other hand, primary samples lacking b-catenin expression were classied as b-catenin.
AML PATIENTS UT7-EPO A B C D E F G H I Jurkat

Cells (0.51 106) were reduced in Laemmli sample buffer and then boiled for 10 min at 1001C. Proteins were resolved on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred to nitrocellulose membrane (Hybond-C Super, Amersham Pharmacia Biotech, Little Chafont, UK). Membranes were blocked in Tris buffer saline TBS-Tween 0.05% with 5% nonfat milk for 1 h, and then incubated with proper antibodies. We purchased monoclonal anti-b-catenin from BD Transduction Laboratories, anti-actin from Sigma (St Louis, MO, USA). All antibodies were diluted 1/1000 in TBSTween 0.05%, except anti-actin that was used 1/10 000. Membranes were blotted overnight at 41C, washed three times, and blotted with appropriate anti-mouse IgG antibody coupled to horseradish peroxydase. Detection was achieved using Pierce Supersignal chemiluminescent substrate.

Western blotting

KG1a

HL60

K562

U937

-catenin Actin -catenin Actin -catenin GAPDH

Clonogenic assays
Leukemic blast primary colony assay was performed as previously described.2223 Briey, 3 105 cells were resuspended in methyl-cellulose semi-solid medium with (PE1 ,
Leukemia

Figure 1 Expression proles of b-catenin in leukemic myeloid cells. b-Catenin is expressed at various protein levels among leukemic cell lines and primary samples obtained from patients, but mRNA is readily expressed suggesting a post-transcriptional regulation. Patients A, F and G are b-catenin, whereas remaining patients are uniformly classied as b-catenin irrespective of protein levels.

b-Catenin and clonogenic capacity in acute myeloid leukemia L Ysebaert et al

1213
Table 1 Initial clinical characteristics of 82 patients with acute myeloid leukemia according to b-catenin expression levels b-Catenin (n 31) Age (median) Gender (%) Male Female Prior MDS (%) %CD34(med./mean:ranges) Karyotype (%) Favorable Intermediate Unfavorable FAB subtype (%) M0 M1 M2 M4 M5 FLT3-ITD mutation (%) Leukocytosis (median) 57 61 39 19 15.5/33.6:1651.2 28/31 patients: 14.3 60.7 25 # 3 35.5 26 i 29 6.5 37 81 200 b-Catenin+ (n 51) 61 49 51 21 61/50.5:35.665.4 45/51 patients: 11 53 36 # 4 22 15 i 35 24 38 50 000 41 0.02 59 NS NS NS NS NS P-value NS NS

64.5 35.5

Fisher exact test (percentages) or MannWhitney (medians) were used to compare variables, and Pvalues are indicated. Flt3-ITD: Flt3 internal tandem duplication, FAB: French-American-British classication, MDS: myelodysplastic syndrome, med.: median values. Karyotypes were classied as favorable (t(8;21) and inv16), intermediate (normal, -Y), and unfavorable (all others, including complex karyotypes).

Table 2

Clonogenic properties of blasts are inuenced by b-catenin expression b-Catenin (n patients) b-Catenin+ (n patients) 500 (n 44) 10550 (n 44) 4100 (n 33) P-value NS 0.048 0.018

PE1: autonomous growth (median number of colonies) PE1+: clonogenic growth after stimulation with 5637 CM (median number of colonies) PE2: replating capacity (median number of colonies)

275 (n 30) 2080 (n 30) 405 (n 16)

Plating efciences (PE) were estimated as described in the Material and methods section, and median number of colonies after 1 (PE1) or 2 (PE2) weeks according to b-catenin status are listed. P-values indicate statistically relevant correlations.

Correlations between b-catenin expression and usual clinico-biologic features in patients

Main characteristics of the population are summarized in Table 1. b-catenin expression was not found to be correlated with CD34 expression in 53 patients (median CD34 cells 61% in b-catenin (n 31) versus 15.5% in b-catenin (n 22) blasts, P 0.12). However, in immature M0M2 subtypes (usually CD34 ), median CD34 expression was 10% (n 13) versus 90% (n 13) in b-catenin and b-catenin , respectively (P 0.024). This correlation was lost in M4M5 leukemias (usually CD34 low), since median CD34 expression was 23 and 29.5%, respectively (P 0.86). As b-catenin was also preferentially expressed in monocytic rather than immature leukemias (M4M5 versus M0M2, P 0.02, Table 1), this result indicates that b-catenin accumulation is not strictly dependent upon CD34 expression.

P ns), but rather PE1 did (median number of colonies 10 550 versus 2080, Po0.05). These results conrm a proliferative advantage under growth factor stimulation for b-catenin blasts. Of note, the expression of the CD34 antigen was not correlated with clonogenic properties of blasts: Median PE1 colonies were 90 versus 50 and PE1 were 9395 versus 4650 (both P ns) according to CD34 expression (chosen cutoff for statistical tests: 20% of CD34 expression). Uppermost, b-catenin is the only variable statistically associated with re-plating capacity (median PE2 was 405 colonies in 16 b-catenin versus 4100 in 33 b-catenin patients, Po0.02, n 49). Again, according to CD34 expression median PE2 colony numbers were 6775 versus 2380 (P 0.45, n 28). Altogether these results indicate that b-catenin expression is a much more reliable predictor of clonogenic potential than the percentage of CD34 cells in the bulk population of blasts.

b-Catenin expression correlates with clonogenic properties of blasts

Results of available clonogenic assays are summarized in Table 2. PE1 (i.e. autonomous growth of blasts) was not inuenced by b-catenin expression (median PE1 colonies 275 versus 500 in b-catenin and b-catenin subjects, respectively,

Clinical outcome is inuenced by b-catenin expression in AML blasts

b-Catenin expression did not inuence CR (data not shown). Expression of b-catenin is associated with both shortened relapse-free survival (RFS, log-rank P 0.012, n 49) and overall survival (OS, log-rank Po0.01, n 82) (Figure 2). In
Leukemia

b-Catenin and clonogenic capacity in acute myeloid leukemia L Ysebaert et al

1214
our cohort, b-catenin is the only factor inuencing RFS in KaplanMeier analysis. Other variables linked to OS (in Kaplan Meier and log-rank tests) are age 460 years (Po0.05), absence of CR (Po0.001), high median plating efciencies (PE1, PE1 and PE2, all with Po0.05), a complex karyotype (Po0.05). In multivariate analysis, b-catenin is an independent variable associated with shortened OS (Po0.05). This correlation between b-catenin and OS is even stronger if we consider the younger (i.e. under the median 62.5 years of age of our cohort) category of patients, with log-rank P 0.004. Altogether, these data suggest that b-catenin is crucial to the pathogenesis of leukemia and the prognosis of patients.

In normal CD34 progenitors b-catenin is strongly expressed but is rapidly lost upon differentiation in cytokines supplemented liquid culture (after 35 days), even if recombinant Wnt3a or Wnt5a were added to culture (our unpublished data and Simon et al.26). The downregulation of b-catenin is phenotypically achieved in CD34 /low38 33 immature progenitors (data not shown), and is also specic to myelomonocytic differentiation. Indeed, under erythroid differentiation conditions, re-expression of high levels of b-catenin in proerythroblasts occurs without any maturation hiatus (our unpublished data). The nding of a loss of b-catenin expression, at a post-transcriptional level, despite Wnt stimulation in normal differentiating progenitors led us to check whether such an antagonistic regulating mechanism could be lost in AML. Wnts are ligands for a co-receptor made by Frizzled and LRP5/6, respectively specically inhibited by FRP-1 and Dkk-1. Some Wnts are pro- (e.g. Wnt3a) and others anti- (e.g. Wnt5a) b-catenin through various couples of coreceptors. b-catenin in leukemic blasts and cell lines was readily stimulated by Wnt3a but not Wnt5a (Figure 3a, representative of 6/6 patients with low/undetectable b-catenin levels), or downregulated by recombinant frizzled related peptide-1 (FRP-1) (in two of three patients) or Dickkopf-1 (Dkk-1) (in one of two patients) (Figure 3b, and data not shown). We next used RT-PCR to check whether differences in b-catenin levels might be linked to the loss of the hematopoietic tumor suppressor Wnt5a (anti-bcatenin ligand) expression in blasts.24 We noticed great variations in Wnt5a transcripts levels in six out of 11 consecutive subjects (Figure 3c), that did not strictly correlate with b-catenin protein expression. Altogether, these results show that the Wnt/b-catenin pathway could be regulated in AML, as opposed to normal myelomonocytic differentiating precursors.

Regulation of b-catenin levels in leukemic myeloid cells differs from normal cells

Figure 2 KaplanMeier relapse-free (RFS) and overall survival (OS) curves in 49 and 82 patients, respectively, according to b-catenin status of leukemic cells. Comparison of survival curves using the logrank test identies b-catenin as an adverse prognostic factor.

Discussion
We aimed to study b-catenin expression in a large cohort of patients to determinate whether it could inuence blast behavior

Patient 1

Patient 2

Patient 3 -catenin Actin

Wnt3a(ng/ml)
0 5 10 20

Wnt3a(g/ml)
0 5 10 20

+ -

+ -

+ -

Wnt3a Wnt5a 1 (+) 2 (+) 3 (-)

U937

KG1 10 (+)

4 (+)

-catenin

Normal CD34+

7 (+)

8 (-)

6 (-)

5 (-)

9 (-)

11 (-)

Patient 4 Patient 5

# patients (-catenin protein level) Wnt5a

Actin + + FRP-1 -catenin GAPDH

Figure 3 b-catenin response prole of leukemic cells upon Wnt ligand stimulation. (a) Leukemic cell lines and patients cells are responsive to Wnt3a (20 ng/ml), but not Wnt 5a (0.3 mg/ml) in terms of b-catenin upregulation, after incubation for 4 h. (b) In b-catenin patients, recombinant FRP-1 (0.1 mg/ml) inhibits b-catenin in the same culture conditions, suggesting that in leukemic cells b-catenin accumulation depends on Wnt stimulatory/inhibitory loops. (c) Expression of Wnt5a mRNAs by reverse transcriptase-polymerase chain reaction (RT-PCR) in normal and leukemic cells. At the mRNA level, Wnt5a is found in normal CD34 progenitors. Yet, six out of 11 patients have lost Wnt5a mRNA expression (anticanonical Wnt), indicating that at least in some AML cells unleashing of canonical Wnt agonists may participate to the accumulation of b-catenin.
Leukemia

b-Catenin and clonogenic capacity in acute myeloid leukemia L Ysebaert et al

in vivo. As already reported by previous publications,2527 we were able to nd b-catenin in the majority of patients (nearly the two-thirds), although at very various levels. For analytical purposes, we chose a simple plus/minus stadication system, which appeared to us as the best way to avoid bias in the classication of patients, avoiding also the choice of a cell type with a reference level of b-catenin (a normal or malignant cell? At which maturation state?). We established a correlation between b-catenin expression and clonogenic growth. Previous reports using clonogenic assays noted the inverse relationship between autonomous growth and the magnitude of response to growth factors, in that cells with low baseline proliferation kinetics are more readily stimulated.28 This is also the case in our work, because b-catenin is correlated with proliferation under stimulation, not at baseline. Moreover, in our series, absence of prior myelodysplastic syndrome is correlated to autonomous growth (not shown), as previously demonstrated.29 Leukemic cell growth in vitro has been correlated to poor OS in our study and in previous publications, and although b-catenin level and clonogenic properties (PE1 and PE2) are correlated, they are both independent prognostic factors for survival. It might be objected that b-catenin does not correlate with autonomous growth, which has been associated with poor prognosis in AML.3031 It should be noticed that translocation products of t(8;21) and t(15;17), that enhance Wnt signaling in leukemic cells, have not been linked to increased autonomous growth and that among AML, those with t(15;17) have been found to have enhanced PE1 rather than PE1.32 The major interest rather comes from the nding of a correlation between b-catenin expression and re-plating efciency, which indicates a putative role of b-catenin in leukemic stem cells (LSCs). Moreover, since b-catenin was the only variable linked to relapse in our cohort, it could mean that a probably small pool of b-catenin-positive immature cells could drive the re-occurrence of the disease, although b-catenin was not correlated with chemoresistance in the bulk AML cells at diagnosis. To understand how Wnt signaling could be deregulated in AML, we sought to determine b-catenin expression in normal hematopoiesis. We conrm that b-catenin expression is rapidly lost upon myelomonocytic differentiation.26,33 Interestingly, our preliminary results show that, depending on the way of maturation, erythroid for example, b-catenin could be reexpressed in CD34 cells (data not shown, and33). We found that in M0M2 AML patients (no M6 AML was tested), b-catenin expression was correlated with the median percentage of CD34 blasts, although this was not the case for M4M5 AML. In unsorted bulk cells, b-catenin expression reects the prevalence of CFU-L more accurately that CD34 expression (our data and8,26,34). The loss of b-catenin in normal precursors coincides with the narrowing of their differentiating capacities (and the loss of CD34 antigen). Thus, the abnormal accumulation of b-catenin in leukemic cells (irrespective of their maturation and CD34 status) may confer clonogenic properties normally restricted to immature precursors. To elucidate the mechanism(s) at stake in maintaining b-catenin in the majority of blasts, we further demonstrated a rapid modulation of b-catenin upon Wnt3a or FRP-1 ligation, indicating that AML blasts are fully responsive to both regulations of b-catenin levels, as opposed to differentiating normal myelomonocytic cells. These data imply that activation/ inactivation of b-catenin by the Wnt pathway is readily achievable in AML cells,35 suggesting that (i) receptors are present, and (ii) downstream Wnt pathways are functional. In the hematopoietic tissue, Wnt5a is a tumor suppressor

antagonizing Wnt signaling through b-catenin.24 In our cohort, we could also nd a differential expression of Wnt5a ligands in some patients (in a way indicating that Wnt signature could be shifted towards oncogenic Wnts at the expense of Wnt5a), but that does not strictly inrm other mechanism(s) to be responsible for b-catenin accumulation in other cases. Whether misexpression of Wnt receptors and/or secreted inhibitors of the pathway could be responsible for variations of b-catenin levels in patients remains to be elucidated. Our data pave the way to developing new therapeutic strategies aimed at inhibiting the Wnt pathway in clonogenic leukemic cells, like micro-environment manipulations (bone marrow stromal cells secrete Wnt ligands to support normal hematopoiesis36), or Wnt-Tcf interactions for example (small molecules inhibitors have been developed37).

1215

Acknowledgements
Research grants and nancial support: LY is supported by a grant from the Fondation de France-Federation Nationale des Centres de Lutte Contre le Cancer. This work was supported by a grant from the Association de Recherche contre le Cancer (contract 3638).

References
1 Pardal R, Clarke MF, Morrison SJ. Applying the priciples of stem cells biology to cancer. Nat Rev Cancer 2003; 3: 895902. 2 Reya T, Morrison S, Clarke M, Weissman I. Stem cells, cancer, and cancer stem cells. Nature 2001; 414: 105111. 3 Polakis P. Wnt signalling and cancer. Genes Dev 2000; 14: 18371851. 4 Giles RH, van Es JH, Clevers H. Caught up in a Wnt storm: Wnt signaling in cancer. Biochim Biophys Acta 2003; 1653: 124. 5 van Es JH, Barker N, Clevers H. You Wnt some, you lose some: oncogenes in the Wnt signaling pathway. Curr Opin Genet Dev 2003; 13: 2833. 6 Moore MAS, Spitzer G, Williams G, Metcalf D, Buckley J. Agar culture studies in 127 cases of untreated acute leukemia: the prognostic value of reclassication of leukemia according to in vitro growth characteristics. Blood 1994; 44: 118. 7 Vincent PC, Sutherland R, Bradley M, Lind D, Gunz FW. Marrow culture studies in acute leukemia at presentation and during remission. Blood 1977; 49: 903912. 8 Del Canizo M, Brufau A, Almeida J, Galende J, Garcia Marcos MA, Mota A et al. In vitro growth in acute myeloblastic leukemia: relationship with other clinico-biological characteristics of the disease. Br J Haematol 1998; 103: 137142. 9 Van den Berg D, Sharma A, Bruno E, Hoffman R. Role of members of the Wnt gene family in human hematopoiesis. Blood 1998; 92: 31893202. 10 Austin T, Solar G, Ziegler F, Liem L, Matthews W. A role for the Wnt gene family in hematopoiesis: expansion of multilineage progenitor cells. Blood 1997; 89: 36243635. 11 Reya T, Duncan AW, Ailles L, Domen J, Scherer D, Willert K et al. A role for Wnt signalling in self-renewal of haematopoietic stem cells. Nature 2003; 23: 409414. 12 Willert K, Brown JD, Danenberg E, Duncan A, Weissman IL, Reya T et al. Wnt proteins are lipid-modied and can act as stem cell growth factors. Nature 2003; 423: 448452. 13 Staal FJ, Clevers HC. WNT signalling and haematopoiesis: a WNTWNT situation. Nat Rev Immunol 2005; 1): 2130. 14 Brandon C, Eisenberg LM, Eisenberg CA. Wnt signaling modulates the diversication of hematopoietic cells. Blood 2000; 96: 41324141. 15 Lu D, Zhao Y, Tawatao R, Cottam H, Sen M, Leoni LM et al. Activation of the Wnt signaling pathway in chronic lymphocytic leukemia. Proc Natl Acad Sci USA 2004; 101: 31183123. 16 Qiang YW, Endo Y, Rubin JS, Rudikoff S. Wnt signaling in B-cell neoplasia. Oncogene 2003; 22: 15361545.
Leukemia

b-Catenin and clonogenic capacity in acute myeloid leukemia L Ysebaert et al

1216
17 Derksen P, Tijn E, Meijer H, Klok MD, MacGillavry MH, van Oers MH et al. Illegitimate Wnt signaling promotes proliferation of multiple myeloma cells. Proc Natl Acad Sci USA 2004; 101: 61226127. 18 Jamieson C, Ailles L, Dylla S, Muijtjens M, Johnes C, Zehnder JL et al. Granulocyte-macrophage pogenitors as candidate leukemic stem cells in blast-crisis CML. N Engl J Med 2004; 351: 657667. 19 Muller-Tidow C, Steffen B, Cauvet T, Tickenbrock L, Ji P, Diederichs S et al. Translocation products in acute myeloid leukemia activate the Wnt signaling pathway in hematopoietic cells. Mol Cell Biol 2004; 24: 28902904. 20 Zheng X, Beissert T, Kukoc-Zivojnov N, Puccetti E, Altschmied J, Strolz C et al. catenin contributes to leukemogenesis induced by AML-associated translocation products by increasing the self-renewal of very primitive progenitor cells. Blood 2004; 103: 35353543. 21 Mc Whirter JR, Neuteboom ST, Wancewicz EV, Monia BP, Downing JR, Murre C. Oncogenic homeodomain transcription factor E2A-Pbx1 activates a novel WNT gene in pre-B acute lymphoblastoid leukemia. Proc Natl Acad Sci USA 1999; 96: 1146411469. 22 Demur C, Muller C, Cassar G, Bousquet C, Laroche M, Laurent G. Acute myeloid leukemia cells with low p-glycoprotein expression and high Rhodamine 123 efux capacity display high clonogenicity. Leukemia 1998; 12: 192199. 23 Laredo J, Demur C, Muller C, Saivin S, Cassar G, Bousquet C et al. Effects of H-7 and staurosporine on proliferation and self-renewal of acute myeloid leukemia progenitors. Leukemia 1993; 7: 813820. 24 Liang H, Chen Q, Coles AH, Anderson SJ, Pihan G, Bradley A et al. Wnt5a inhibits B cell proliferation and functions as a tumor suppressor in hematopoietic tissue. Cancer Cell 2003; 4: 349360. 25 Chung EJ, Hwang SG, Nguyen PM, Lee S, Kim JS, Kim JW et al. Regulation of leukemic cell adhesion, proliferation, and survival by b-catenin. Blood 2002; 100: 982990. 26 Simon M, Grandage V, Linch D, Khwaja A. Constitutive activation of the Wnt/b-catenin pathway in acute myeloid leukemia. Oncogene 2005; 24: 24102420. 27 Tickenbrock L, Schwable J, Wiedehage M, Steffen B, Sargin B, Choudhary C et al. Flt3 tandem duplication mutations cooperate with Wnt signaling in leukemic signal transduction. Blood 2005; 105: 36993706. 28 Norgaard JM, Langkjer S, Palshof T, Clausen N, Pedersen b, Hokland P. Relation of blast cell survival and proliferation to chemotherapy resistance in AML. Br J Haematol 1996; 93: 888897. 29 Richert-Boe KE, Bagby GC. In vitro hematopoiesis in myelodysplasia: liquid and soft-gel culture studies. Hematol Oncol Clin North Am 1992; 6: 543556. 30 Hunter A, Rogers S, Roberts I, Barrett AJ, Russel N. Autonomous growth of blast cells is associated with reduced survival in acute myeloblastic leukemia. Blood 1993; 82: 899903. 31 Lowenberg B, Van Putten W, Touw I, Delwel R, Santini V. Autonomous proliferation of leukemic cells in vitro as a determinant of prognosis in adult acute myeloid leukemia. N Engl J Med 1993; 328: 614619. 32 Ohler L, Berer A, Aletaha D, Kabrna E, Heinze G, Streubel B et al. Cytogenetic risk groups in acute myeloblastic leukemia differ greatly in their semi-solid colony growth. Br J Haematol 2001; 113: 120125. 33 Serinsoz E, Neusch M, Busche G, Wasielewski R, Kreipe H, Bock O. Aberrant expression of beta-catenin discriminates acute myeloid leukaemia from acute lymphoblastic leukaemia. Br J Haematol 2004; 126: 313319. 34 Tepstra W, Prins A, Ploemacher RE, Wogum BW, Wagemaker G, Lowenberg B et al. Long-term leukemic-initiating capacity of a CD34- subpopulation of acute myeloid leukemia. Blood 1996; 87: 21872193. 35 Baco A, Liu G, Goldin L, Harris V, Aaronson SA. An autocrine mechanism for constitutive Wnt pathway activation in human cancer cells. Cancer cell 2004; 6: 497506. 36 Yamane T, Kunisada T, Tsukamoto H, Yamazaki H, Niwa H, Takada S et al. signaling regulates hemopoiesis through stromal cells. J Immunol 2001; 167: 765772. 37 Lepourcelet M, Chen YN, France DS, Wang H, Crews P, Petersen F et al. antagonists of the oncogenic Tcf/beta-catenin protein complex. Cancer Cell 2004; 5: 91102.

Leukemia

A critical role for Lyn in Acute Myeloid Leukemia Cdric Dos Santos,1 Ccile Demur,1,3 Valrie Bardet,4 Nais Prade-Houdellier,1 Bernard Payrastre1* and Christian Rcher1,2*

Inserm, U563, Centre de Physiopathologie de Toulouse Purpan et Universit Paul Sabatier, Dpartement dOncogense, Signalisation et Innovation thrapeutique, Toulouse, France ; Universit Toulouse III Paul Sabatier, IFR30, Toulouse, France ;1 CHU Toulouse, Hpital Purpan, Service dHmatologie, Toulouse, France ;2 CHU Toulouse, Hpital Purpan, Laboratoire dHmatologie, Toulouse, France;3 Universit Paris Descartes, Institut Cochin (Inserm U567, CNRS-UMR 8104), Service dHmatologie Biologique, AP-HP, Hpital Cochin, Paris, France.4

Reprints: Dr Christian Rcher, Service dHmatologie, CHU Purpan, place du Dr Baylac, 31059 Toulouse cedex, France ; phone :+33561772078 ; fax : +33561777541 ; e-mail : recher.c@chu-toulouse.fr. Running head: Lyn activity in AML Word count: 4309 Abstract: 198 Scientific category: Neoplasia C.D.S participated in designing, performing the research and analyzed data; C.D. centralized the cytological review and performed the research; and V.B and N.P.H performed the research, B.P and C.R. controlled, analyzed data, and wrote the paper. All authors checked the final version of the manuscript. BP and CR contributed equally to this study. The online version of the article contains a supplement. Supported by grants from La Ligue Nationale contre le Cancer, the Association pour la Recherche sur le Cancer (contract N 3122), the Association Laurette Fugain and the Institut National du Cancer.

Abstract Receptor or non-receptor tyrosine kinases (TKs) are known to play an important role in leukemogenesis. Here we studied the level of protein tyrosine phosphorylations in a series of fresh AML samples and evaluated the effect of TK inhibitors. Compared to normal haematopoietic progenitors, a high level of tyrosine phosphorylation was detected in most AML samples. The Src family kinases (SFKs) appeared constitutively activated in most cases including in the CD34+ CD38- CD123+ compartment as revealed by the level of phosphorylated tyrosine 416. Lyn was the major SFK family member expressed in an active form in AML cells where it was abnormally distributed throughout the plasma membrane and the cytosol as opposed to normal haematopoietic progenitors. The SFK inhibitor, PP2, strongly reduced the global level of tyrosine phosphorylations, inhibited cell proliferation and induced apoptosis in patient samples without affecting normal granulo-monocytic colony forming units. Moreover, silencing Lyn expression by small interfering RNA in primary AML cells strongly inhibited proliferation. Interestingly, a link between Lyn and the mTOR pathway was observed as PP2 and a Lyn knockdown both affected the phosphorylation of mTOR targets without inhibiting Akt phosphorylation. Lyn should be considered as a novel pharmacological target for AML therapy.

Introduction Acute Myeloid Leukemia (AML) is characterized by an accumulation of neoplastic myeloid precursors in the bone marrow and blood. The processes of neoplastic transformation are thought to arise at the haematopoietic stem cell or committed myeloid progenitor level, leading to the genesis of leukemic haematopoiesis sharing organizational similarities with its normal counterpart.1,2 Despite an apparent uniformity, AML cells are classified in different compartments. The less mature compartment includes a small fraction of leukemic stem cells (LSC) that are mostly quiescent but capable of self-renewal or can be committed into proliferation/differentiation. Another compartment contains the more mature and highly cycling progenitors (CFU-L) that are incapable of self-renewal but largely committed to proliferation and limited differentiation. CFU-L gives rise to a large population of differentiated leukemic cells (blast cells), which are arrested at a given stage of granulomonocytic differentiation and represent the majority of leukemic cells in the blood and marrow of AML patients. From a clinical point of view, AML is a heterogeneous disease with marked differences in response to therapy, occurrence of relapses and overall survival according to prognostic factors such as age, cytogenetic or molecular abnormalities.3 Many different genetic defects are though to support the heterogeneity of this disease. Alterations in self-renewal, differentiation and survival observed in AML cells are the consequence of two types of cooperating mutations. Mutations affecting transcription factors (class II mutations) lead to impaired differentiation whereas mutations of receptor tyrosine kinases (RTKs) such as FLT3, c-KIT or downstream effectors like Ras (class I mutations) enhance cell survival and proliferation.4 The molecular characterization of recurrent translocations and the development of murine models of the disease have established the common idea that these two types of mutations are sufficient to generate the leukemic phenotype. Activating mutations of RTKs have been identified in only 50% of all AMLs.

Therefore, one can expect the involvement of other deregulated RTKs and/or nonreceptor protein tyrosine kinases (TKs) acting as key activators of intracellular signal-transduction pathways. Indeed, cell survival and proliferation pathways including MAPK,

Phosphoinositide 3-kinase (PI3K)/Akt, mTOR, NF-kB or STATs are deregulated in most, if not all, AML cases.5-13 Here, we investigated the profile of tyrosine phosphorylations in a large panel of fresh AML samples. We observed a high level of tyrosine phosphorylation in AML samples as compared to normal haematopoietic progenitor cells (HPCs). Tyrosine kinase inhibitors, particularly the Src family kinase (SFKs) inhibitor, PP2, strongly affected the level of global protein tyrosine phosphorylation. Moreover, PP2 induced the inhibition of cell proliferation and apoptosis in AML cell lines and patient samples without affecting colony-forming unitgranulocyte macrophage growth. SFKs appeared constitutively activated in leukemic cell lines, in AML samples as well as in the immature compartment of the leukemic clone (CD34+CD38-CD123+). Among SFKs, Lyn was consistently expressed at a high level and constitutively activated in leukemic cells. Accordingly, the specific down regulation of Lyn expression by small interfering RNA (siRNA) inhibited the clonogenic properties of AML cells. Moreover, PP2 and Lyn siRNA induced the inhibition of S6-ribosomal protein (RPS6) and 4E-BP1 phosphorylation, demonstrating a novel link between Lyn and the mTOR pathway in primary AML cells. These results indicate that Lyn is important for survival and proliferation of AML cells and represents a novel potential pharmacological target for antileukemic drugs.

Material and methods Cells and reagents AML samples were obtained from patients diagnosed at the Department of Hematology, Toulouse University Medical Center (France), after informed consent. The institutional review board of the Hpital Purpan, Toulouse, France, approved the study. AML cells and normal bone marrow CD34+ HPCs were isolated and processed as previously described.14 To assess the stimulation of the global phosphotyrosine profile, 5x105 CD34+ HPCs were incubated in IMDM 10% BIT (Bovine Serum Albumin, Insulin, Transferrin) with or without Granulocyte-Macrophage Colony-Stimulating Factor (GM-CSF, 10 ng/mL), Stem Cell Factor (SCF, 100 ng/mL), and Interleukin-3 (IL3, 1 IU/mL) for 5, 15, 20 minutes or 4 hours, and then processed for western blot analysis (see western blotting). The human leukemic cell lines KG1, U937 and UT7-EPO were purchased from American Type Culture Collection (ATCC; Rockville, MD) and cultured in IMDM containing 20% FCS (KG1); RPMI 10% FCS (U937) and MEM 10% FCS and 1 UI/L Erythropoietin (UT7-EPO). All AML cell lines were incubated in a humidified CO2 incubator (5% CO2, 37C). Rapamycin and LY294002 were purchased from Sigma-Aldrich (St Louis, MO); PP2 [4-amino-5-(4-chlorophenyl)-7(tbutyl)pyrazolo[3,4-D]pyramidine], AG1296 and AG490 were purchased from Calbiochem (San Diego, CA). All these inhibitors were dissolved in DMSO, and stored at -20C. Antibodies used included: antiphospho-Akt (Thr308 and Ser473), antiphospho-Src (Y416), antiphospho-p70S6K (Thr389), antiphospho4E-BP1 (Thr70), anti-phospho-RPS6

(Ser235/236), anti-p70S6K, anti4E-BP1, anti-S6-ribosomal protein, anti-caspase 3, horseradish peroxidaseconjugated secondary antibodies against mouse and rabbit immunoglobulins from Cell Signaling (Beverly, MA); anti-phospho Erk1/2 from SigmaAldrich; anti-Akt, anti-Fgr, anti-Fyn, anti-Hck, anti-Lck, anti-Lyn, anti-Src and anti-PARP

from Santa Cruz Biotechnology (Santa Cruz, CA) and anti-phosphotyrosine (mAb; clone 4G10) from Upstate Biotechnology (Charlottesville, VA).

Western blotting and immunoprecipitation Western blotting was performed as previously described.14 For immunoprecipitation, 107 fresh cells were lysed for 30 minutes at 4C in 800 L of a lysis buffer containing 20 mM Tris/HCl pH 7,4, 150 mM NaCl, 1 mM EDTA, 0,8% Triton X-100, 4 mM NaF, 1 mM orthovanadate and a protease cocktail inhibitor. After centrifugation, the supernatants were incubated for 1 hour at 4C with 3 g of Lyn antibody, followed by the addition for 1 hour of Protein Asepharose. After 3 washes with the same buffer without Triton X-100, the pellets were lysed in a Laemlis buffer, loaded onto SDS-PAGE and immunostained with the appropriate antibody. Lyn and p-Src antibodies were biotinylated using the BiotinTag Micro

Biotinylation Kit (Sigma-Aldrich) and detected using a horseradish peroxidase coupled to streptavidin (Pierce, Rockford, IL).

Small Interfering RNA experiments Fresh AML cells, KG1 and U937 were transfected using the Amaxa nucleofection technology (Amaxa, Koeln, Germany). Briefly, cells were resuspended in Amaxa solution kit R (KG1) and V (fresh samples and U937) following the Amaxa guidelines for cell lines. Leukemic cells (2 x 106) in 100 L of solution kit V or R were mixed with 3g siRNA smartpool Lyn or with 3g siRNA control both for cell lines and primary AML samples (Dharmacon Inc., Lafayette, CO) and immediately nucleofected with an Amaxa Nucleofector apparatus (program V-01 for both cell lines, program U-15 for primary AML cells). Cells were then immediately transferred into wells containing 37C prewarmed culture medium in 6-well

plates. After transfection, cells were cultured from 24 to 72 h before analyzing by Western Blot.

Flow cytometry analysis Fresh leukemic cells and KG1 (2x105) were incubated for 1 hour with or without PP2, then fixed and permeabilized with intraprep kit (Beckman coulter, Miami, FL) according to the manufacturers recommendations. Cells were then incubated for 1 hour at +4C in the dark with antiphospho-Src (Y416), rabbit non-relevant IgG (Santa Cruz) or no antibody. Samples were then washed twice with PBS and pellets were incubated for 30 minutes in the dark with the secondary Goat anti Rabbit FITC antibody (BD Biosciences). All samples were analyzed by Epics XL (Beckman Coulter, Villepinte, France) and results were expressed as the ratio between mean fluorescence intensity of the stained sample and mean fluorescence intensity of the control. For several experiments on AML samples, a double immunostaining procedure combining Pe-Cy5 conjugated anti-CD34 (Beckman coulter) surface staining and anti phospho-Src (Y416) intracellular staining was also conducted with the same results. Flow cytometric detection of the immature compartment of the leukemic clone (CD34+CD38CD123+ cells) has been described elsewhere.15

Apoptosis Fresh AML cells (2x106) were incubated in 96-well flat-bottomed plates with or without 10 M of PP2. After incubation, cells were washed twice with cold PBS (without Ca2+ and

Mg2+), followed by addition of 5 L of propidium iodide and 5 L of annexin V-FITC. Cells were also incubated for 15 minutes at room temperature in the dark. The percentage of apoptotic cells was assessed by flow cytometry using an Epics XL (Beckman Coulter).

Clonogenic assays Sensitivity of leukemic progenitors to PP2. The percentage of leukemic cells after Ficoll separation ranged between 60% and 100% (median; 92%). AML cells were adjusted to a final concentration of 1x105 cells/mL and grown in H4230 Stem Cell Technologies methyl cellulose medium (Stem Cell Technologies, Vancouver, BC) supplemented with 10% 5637conditionned medium (CM) as a stimulant (5637 is a bladder carcinoma cell line secreting GM-CSF, G-CSF, IL1beta, M-CSF, and SCF) 16 and increasing concentrations of PP2 (0, 0, 5, 1, 2, 5, 10, 20 M). The cells were then plated in 35 mm Petri dishes in duplicate and incubated for 7 days in a humidified CO2 incubator (5% CO2, 37C). Leukemic colonies (more than 20 cells) and clusters (more than 5 cells) were scored under an inverted microscope at day 7 and were uniformly named CFU-L (Colony Forming Unit-Leukemia). In each case, the viability and the leukemic nature of CFU-L were confirmed by a blue trypan exclusion assay and by morphological analysis after Giemsa staining. We also performed karyotypic analysis in some samples and, as expected, in each case, colony-forming cells displayed identical cytogenetic abnormalities to the primary leukemic cells. Colony-forming assays of human bone marrow CD34+ cells. Fresh CD34+ human bone marrow cells were washed twice in IMDM medium with 10% FCS and resuspended at a concentration of 7 x 103/mL in Stem Cell Technologies methyl cellulose complete media: H4230 supplemented with 10% 5637-CM (CFU-GM growth); H4435 (CFU-M growth) and H4535 (BFU-E growth). PP2 was added at a concentration of 0, 2, 5 and 10 M. The cells were then plated in 35-mm Petri dishes and incubated in a humidified CO2 incubator (5% CO2, 37C) for 14 days. Colonies (more than 50 cells) were then scored under an inverted microscope.

Confocal microscopy AML primary cells, KG1 and U937 (2x105) were plated on poly-L-lysine pre-coated slides, fixed with 3.7% (w/v) paraformaldehyde for 30 minutes, washed twice with PBS and then permeabilized by 0.2% (v/v) Triton X-100 in BSA-PBS (w/v) for 5 minutes. Slides were incubated for 30 minutes with 3% BSA-PBS for aspecific binding and then incubated for 30 minutes again with primary antibodies diluted in 3% BSA-PBS (1/500 for Lyn and 1/100 for p-Src Y416). After 3 washes in 3% BSA-PBS, secondary antibodies (anti-mouse FITC antibody and anti-rabbit Alexa Fluor594) were added for 30 minutes. After several washes in PBS, the coverslips were mounted with Moewiol and all the preparations were analyzed by confocal laser scanning using a Zeiss LSM510 equipped with a 63 x objective.

Results

The level of protein tyrosine phosphorylation is high in AML and sensitive to SFK inhibition Abnormal activation of tyrosine kinases and of the signaling pathways they control is thought to play a critical role in leukemogenesis. To assess the role of tyrosine kinases in AML cells, we first examined the level of tyrosine phosphorylations in normal HPCs, AML samples (n=38, Table I) and leukemic cell lines (KG1, U937, UT7-Epo) by immunoblotting with an anti-phosphotyrosine antibody. In normal quiescent CD34+ HPCs, the level of tyrosine phosphorylation was low but increased upon stimulation with a cocktail of cytokines (SCF, IL3, GM-CSF). These phosphorylations increased rapidly and returned to the basal level after 4 hours of stimulation. Although somewhat heterogeneous, the tyrosine phosphorylation levels were consistently high in AML samples (Figure 1A). It is noteworthy that the tyrosine phosphorylation profiles appeared more homogeneous when assessed according to their cytogenetic or molecular background (Supplementary Figure 1). Strikingly, we observed a strong signal of tyrosine phosphorylation of proteins of about 55-60 kDa. This molecular weight may correspond to SFKs, which are known to be expressed in AML cells.17 Moreover, the SFK inhibitor PP2 dramatically reduced the degree of tyrosine phosphorylation in most samples tested (15 out of 20), suggesting that these kinases play an important role in AML (Figure 1B).

SFKs are constitutively phosphorylated on Y416 both in leukemic bulk and in immature CD34+ CD38- CD123+ leukemic cells The phosphospecific antibody recognizing the activated, Tyr416-phosphorylated form of SFK members (anti-pSrc Y416), was used to monitor the activation state of these kinases by

10

immunoblotting and flow cytometry. The level of activation of SFKs was assessed in normal CD34+, in the KG1 cell line and in a series of 53 AML samples (Figure 2A). Normal CD34+ cells displayed a low level of phosphorylation on Y416 but this phosphorylation increased after 5 and 20 minutes of cytokine stimulation. Contrasting with normal cells, SFKs were found to be highly phosphorylated on Y416 in all AML samples tested as well as in KG1 cells. Flow cytometry allows the detection of the immature compartment of the leukemic clone which expresses the CD34+CD38-CD123+ phenotype15. This sub-compartment is thought to be enriched in leukemic stem cells.7,18,19 Interestingly, CD34+CD38-CD123+ leukemic cells displayed a high level of SFK phosphorylation on Y416 that was inhibited by PP2 (Figure 2 B and 2C). Indeed, the ratio between the mean fluorescence intensity (MFI) of the stained cells and the MFI of the isotypic control (RFI) was 1.5 0.3, 4.1 2.3 and 4.6 2.5 in normal CD34+ HPCs, leukemic bulk and CD34+CD38-CD123+ cells (mean of 5 AML patients tested), respectively. After one hour of incubation with 10 M PP2, the mean RFIs decreased in the leukemic bulk and CD34+CD38-CD123+ cells (Figure 2C) and in KG1 cells (not shown). Thus, SFKs are phosphorylated on Y416 activated in the leukemic bulk as well as in early leukemic progenitors but not in their normal counterpart.

Lyn is highly expressed and constitutively phosphorylated on Y416 in AML cells In order to identify the members of SFKs implicated in AML, we analyzed the level of expression of the various members of the family. In a large panel of fresh samples (n=33), we found that Lyn, Hck and Fgr were detected in 100%, 78% and 66% of cases, respectively. Conversely, Fyn was weakly detected in most AML samples and Lck was hardly or not detectable in the majority of cases (Figure 3A and Supplementary Figure 2A). It is noteworthy that the major bands detected by the anti-pSrc Y416 antibody in KG1 and in patient samples

11

shown in Figure 2A, matched the two Lyn kinase isoforms (p53 and p56). When the nitrocellulose membrane shown in Figure 1A was stripped and reprobed for Lyn, the signal obtained matched the major band of 50/60 kDa observed with the anti-phosphotyrosine antibody (Figure 1A). Depletion of Lyn from the lysate of U937 and KG1 cells strongly decreased the 50/60 kDa phosphotyrosine band indicating that Lyn accounts for the major part of this signal (Supplementary Figure 2B). Moreover, as shown by the immunoprecipitation of Lyn and immunoblotting with the anti-pSrc Y416, Lyn was phosphorylated on the surrogate marker of its active form in KG1 cells as well as in fresh AML samples (Figure 3B). Accordingly, specific knockdown of Lyn expression using siRNA induced a dramatic reduction of the anti-pSrc Y416 signal in KG1 cells, U937 and fresh AML cells (Figure 3C). Furthermore, an in vitro kinase assay based on the auto-phosphorylation of

immunoprecipitated Lyn has revealed a highly active kinase in leukemic cells compared to normal cells (Supplementary Figure 2C). Lyn distribution was analyzed in normal CD34+ HPCs, KG1 cells, U937 and AML samples by confocal microscopy. The distribution of Lyn in normal CD34+ HPCs was restricted to distinct points at the plasma membrane and the signal obtained with the anti-pSrc Y416 was very weak (Figure 3D). In contrast, Lyn appeared to be distributed over the entire plasma membrane and also in the cytoplasm where it colocalized with the anti-pSrc Y416 staining (Figure 3D).

Anti-leukemic activity of PP2 and Lyn siRNA To further characterize the role of SFKs in AML cells, we analyzed the effect of the SFK inhibitor PP2 on survival and proliferation of KG1 cells and AML samples. PP2 inhibited KG1 cell growth in a time and dose dependent manner both in clonogenic assays and in liquid culture and this effect was mainly due to a cell cycle arrest in G1 (Supplementary Figure 3).

12

Since fresh AML cells usually hardly proliferate in liquid culture, we used a clonogenic assay in methylcellulose conditioned with haematopoietic growth factors to study the effect of PP2 on samples from 8 AML patients. PP2 inhibited the capacity of AML cells to generate CFUL, in a dose dependent manner, with an IC50 below 10 M in most cases (5/8) (Figure 4A, left panel). We then studied the effect of PP2 on the capacity of normal CD34+ HPCs to generate granulo-monocytic (CFU-GM), erythroid (BFU-E) and monocytic (CFU-M) colonies. No significant change in the number of colony forming CFU-GM was observed upon PP2 treatment. The number of BFU-E and CFU-M was significantly decreased upon treatment with the highest concentration of PP2 (Figure 4A, right panel). Altogether, these results demonstrate that, in most cases, PP2 targets AML progenitors at concentrations below 10 M while having no effect on normal granulo-monocytic progenitors. As shown in Figure 4B, incubation of fresh AML samples with 10 M PP2 for 24 hours significantly induced apoptosis as assessed by annexin V binding (46 %, PP2 vs 18%, control), PARP cleavage and decrease in full length caspase 3. As Lyn appeared to be a major SFK in AML we analyzed the effect of a specific decrease of its expression in fresh AML samples using siRNA. Figure 4C shows that the knockdown of Lyn strongly reduced the clonogenic properties of AML cells.

The SFK inhibitor PP2 but not AG1296 or AG490 inhibits the phosphorylation of mTOR targets in the KG1 cell line and AML samples. We have recently demonstrated that the mTOR/4E-BP1/p70S6K pathway is constitutively activated in AML cells but not in their normal counterpart.14 Rapamycin, a selective inhibitor of mTOR, inhibits the capacity of AML cells to generate CFU-L without affecting the growth of normal CD34+ HPCs. However, the mechanisms of activation of the mTOR/4E-

13

BP1/p70S6K pathway in AML remain poorly understood. Since the effect of PP2 on both leukemic cells and normal HPCs stickingly mimicked rapamycin activity, we hypothesized a putative regulatory role of SFKs on mTOR activity in AML cells. We first treated KG1 cells, which display a high level of mTOR activation, with PP2 and other tyrosine kinase inhibitors including AG1296 (targeting c-Kit, PDGFR and FLT3) and AG490 (targeting JAK2). As expected, rapamycin inhibited the phosphorylation of mTOR targets. Interestingly, PP2 (10 M) also inhibited the phosphorylation of p70S6K and 4E-BP1, whereas AG1296 and AG490 had no effect (Figure 5A). Moreover, a 1h treatment of fresh AML samples with PP2 consistently inhibited the phosphorylation of p70S6K and 4E-BP1 without affecting their expression level (Figure 5B). These results suggested that member(s) of the SFKs could play a role in regulating the mTOR pathway in AML cells. Consistent with the results obtained in the leukemic bulk, the high level of SFKs activation in CD34+CD38-CD123+ cells correlated with an elevated phosphorylation of RPS6. RPS6 phosphorylation was also inhibited by PP2 treatment in this immature sub-compartment (not shown). Interestingly, PP2 did not affect Akt phosphorylation suggesting that SFKs are acting either downstream of this kinase in the mTOR pathway or in an alternative regulatory pathway. It is noteworthy that the PI 3-kinase inhibitor LY294002 had no significant effect on the SFKs Y416 phosphorylation suggesting that these pathways are distinct (Figure 5C). In order to investigate whether Lyn is involved upstream of mTOR, the level of phosphorylation of mTOR targets was assessed after silencing Lyn with siRNA in KG1 and U937 cell lines (Figure 6A) and in AML samples (Figure 6B). The reduction of Lyn expression induced a significant decrease in the level of phosphorylation of 4E-BP1 and RPS6 as compared to control siRNA. Again, Lyn knockdown did not influence the phosphorylation of Akt both in cell lines and in fresh AML samples. As a control, the activation of another unrelated key signaling protein, Erk1/2, was not affected by Lyn knockdown.

14

Discussion In this study, we show that AML cell lines and freshly isolated cells from AML patients exhibit a high level of tyrosine phosphorylation as compared to their normal counterpart. Strikingly, all AMLs displayed high levels of SFKs activation regardless of their cytogenetic or molecular background. SFK activation was also observed in the CD34+CD38CD123+ compartment of the leukemic clone, indicating that SFKs are activated in the subcompartment enriched in leukemic stem cells. The SFKs inhibitor PP2 efficiently inhibited cell proliferation and survival both at the leukemic bulk and clonogenic levels. Among the SFKs, we show in a large series of AML cases, that activated Lyn is consistently expressed at a high level. Two other members of the family, Hck and Fgr, are also strongly expressed in a significant proportion of AML patients. We focused our interest on Lyn as it appeared to be responsible for the main SFK activity in leukemic cells as shown by the silencing experiments. Lyn was phosphorylated on tyrosine residue 416 suggesting that it was under an activated form although we cannot exclude that the phosphorylation of other sites also has an impact on its activity.20 However, an in vitro kinase assay based on the autophosphorylation of immunoprecipitated Lyn has revealed a highly active kinase in leukemic cells as compared to normal cells. Consistent with our observations, previous reports indicated that Lyn was activated both in a small series of AML samples and in cell lines.17,21 Moreover, we show that in AML cells, Lyn is distributed over the entire plasma membrane and in the cytoplasm possibly due to association with internal membranes as previously described in chronic lymphocytic leukemia cells.22 Conversely, Lyn is concentrated in several distinct points across the membrane in normal CD34+ HPCs and the level of SFKs activation is low in these cells. In normal cells, Lyn is known to concentrate in lipid rafts where it is tightly regulated by the C-terminal Src kinase-binding protein (Cbp)-CSK complex, which phosphorylates tyr507 resulting in its inactivation.23-25 Thus, mislocalized Lyn could escape

15

the Cbp-CSK-mediated negative regulation in leukemic cells. Loss of the negative regulation of SFKs could also be explained by a possible defect in the activity of phosphatases such as SHP1 and SHP2.26-28 Other regulators of the level of expression of SFKs such as the ubiquitine ligase Cbl or the suppressor of cytokine signalling (SOCS) proteins may also contribute to the deregulation of SFKs in AML.29 Indeed, inactivating mutations of c-Cbl and Cbl-b have recently been shown in AML patient cells contributing to leukemogenesis through RTK activation.30,31 Interestingly, Lyn is considered as a negative regulator of normal myeloid haematopoiesis, except for terminal maturation of erythroblasts, and can have both positive and negative roles in B cells.25,32-35 Thus, according to the cell context, Lyn can have opposite effects on the regulation of signaling pathways. The high level of Lyn expression and its aberrant activation in leukemic cells possibly causes it to switch from being a negative to a positive regulator of cell proliferation, as recently proposed in Philadelphia chromosomepositive acute lymphoblastic leukemia.36,37 Consistent with this idea, PP2 has no significant effect on normal CFU-GM proliferation and differentiation, whereas it strongly blocks the clonogenic properties of AML cells in a dose dependent manner. Importantly, it has been shown that specific down-regulation of Lyn had no effect on the survival of normal HPCs.37 In contrast, the specific knockdown of Lyn expression by siRNA significantly inhibited the clonogenic properties of fresh AML cells, comparable to the effect of the pan SFK inhibitor PP2. Similar results have been obtained in haematopoietic cell lines expressing mutated RTK such as FLT3.21,38 Overall, these results strongly suggest an important role for Lyn in AML biology. One critical pathway overactivated in most AML cases is the mTOR/p70S6K/4E-BP1 pathway. In this study, we show that Lyn plays a role in the activation of this pathway. Indeed, the inhibition of SFKs by PP2 strongly affected the phosphorylation of p70S6K and

16

4E-BP1 both in cell lines and in fresh AML samples. Moreover, the treatment with Lyn siRNA decreased the phosphorylation of mTOR targets. One could hypothesize that Lyn regulates the PI3K/Akt pathway upstream of mTOR/p70S6K/4E-BP1 as it was previously shown that Lyn activates PI3K/Akt in response to G-CSF in non leukemic haematopoietic cells.39,40 However, both PP2 and Lyn siRNA efficiently inhibited the phosphorylation of mTOR targets without affecting Akt phosphorylation. This result suggests that Akt signaling is not sufficient for mTOR activation in AML. In agreement, a recent report indicated that Lyn silencing in FLT3-ITD expressing 32D cells has no effect on Akt phosphorylation.38 It is also noteworthy that the selective inhibition of PI 3-kinase abolishes Akt activation in a large proportion of AML samples without affecting mTOR in most cases.12,41,42 Thus, alternative mechanisms other than the well established PI3K/Akt pathway may link TKs, particularly Lyn, to mTOR in AML. In this respect, it is noteworthy that a recent report suggests that another non-receptor tyrosine kinase, Syk, can activate the mTOR pathway independently of Akt in follicular lymphoma cells.43 How tyrosine kinases can regulate the mTOR pathway is still unknown. However, consensus tyrosine phosphorylation sites for SFKs (EEXIYGEIEA) are found in several upstream regulators of mTOR including TSC1/2 and Rheb and in mTOR itself as shown through sequence analysis (NetPhos 2.0 Server) suggesting possible direct regulations. As in addition to growth factors or oncogenes mTOR integrates signals from nutrients and energy status we cannot exclude a possible role for Lyn in regulating one of these mechanisms. Of note, Lyn did not appear to regulate the

ERK/MAPK pathway, which is frequently activated in AML.5 This result is consistent with another study showing that Lyn downregulation induced a reduction of STAT5 phosphorylation without affecting MAPK and Shc.38 Our study demonstrates that Lyn is highly expressed and activated in AML cells, including leukemic progenitors. Lyn controls cell survival and proliferation as demonstrated

17

by the anti-leukemic activity of both SFK inhibitors and Lyn siRNA. Surprisingly, ablation of Lyn expression in primary cells uncovered a novel link between this kinase and the activation of the mTOR pathway. Altogether, these results support the use of SFK and Lyn inhibitors for new therapeutic approaches in AML.

Acknowledgments: We thank Sophie Allart and Fatima LFaqihi from the Imaging and flow cytometry Core Facility of IFR 30; Monique Larroche and Nicole Lhermie for technical assistance and Drs S Manenti, C Racaud-Sultan and H Tronchre for helpful discussions.

18

References 1. Bonnet D, Dick JE. Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell. Nat Med. 1997; 3: 730-737. 2. Passegue E, Jamieson CH, Ailles LE, Weissman IL. Normal and leukemic hematopoiesis: are leukemias a stem cell disorder or a reacquisition of stem cell characteristics? Proc Natl Acad Sci U S A. 2003; 100 (Suppl 1): 11842-11849. 3. Tallman MS, Gilliland DG, Rowe JM. Drug therapy for acute myeloid leukemia. Blood. 2005; 106: 1154-1163. 4. Kelly LM, Gilliland DG. Genetics of myeloid leukemias. Annu Rev Genomics Hum Genet. 2002 ; 3:179-98. 5. Milella M, Kornblau SM, Estrov Z, et al. Therapeutic targeting of the MEK/MAPK signal transduction module in acute myeloid leukemia. J Clin Invest. 2001; 108: 851-859. 6. Hayakawa F, Towatari M, Kiyoi H, et al. Tandem-duplicated Flt3 constitutively activates STAT5 and MAP kinase and introduces autonomous cell growth in IL-3-dependent cell lines. Oncogene. 2000; 19: 624-631. 7. Guzman ML, Neering, SJ, Upchurch D, et al. Nuclear factor-kappaB is constitutively activated in primitive human acute myelogenous leukemia cells. Blood. 2001; 9: 2301-2307. 8. Min YH, Eom JI, Cheong JW, et al. Constitutive phosphorylation of Akt/PKB protein in acute myeloid leukemia: its significance as a prognostic variable. Leukemia. 2003; 17: 995997. 9. Xu Q, Simpson SE, Scialla TJ, Bagg A, Carroll M. Survival of acute myeloid leukemia cells requires PI3 kinase activation. Blood. 2003; 102: 972-980. 10. Birkenkamp KU, Geugien M, Schepers H, Westra J, Lemmink HH, Vellenga E. Constitutive NF-kappaB DNA-binding activity in AML is frequently mediated by a Ras/PI3K/PKB-dependent pathway. Leukemia. 2004; 18: 103-112. 11. Grandage VL, Gale RE, Linch DC, Khwaja A. PI3-kinase/Akt is constitutively active in primary acute myeloid leukaemia cells and regulates survival and chemoresistance via NFkappaB, Mapkinase and p53 pathways. Leukemia. 2005; 19: 586-594. 12. Sujobert P, Bardet V, Cornillet-Lefebvre P, et al. Essential role for the p110{delta} isoform in phosphoinositide 3-kinase activation and cell proliferation in acute myeloid leukemia. Blood. 2005; 106: 1063-1066. 13. Recher C, Dos Santos C, Demur C, Payrastre B. mTOR, a new therapeutic target in acute myeloid leukemia. Cell Cycle. 2005; 4: 1540-1549.

19

14. Recher C, Beyne-Rauzy O, Demur C, et al. Antileukemic activity of rapamycin in acute myeloid leukemia. Blood. 2005; 105: 2527-2534. 15. Bardet V, Tamburini J, Ifrah N, et al. Single cell analysis of phosphoinositide 3kinase/Akt and ERK activation in acute myeloid leukemia by flow cytometry. Haematologica. 2006; 91: 757-764. 16. Quentmeier H, Zaborski M, Drexler HG. The human bladder carcinoma cell line 5637 constitutively secretes functional cytokines. Leuk Res. 1997; 21: 343-350. 17. Roginskaya V, Zuo S, Caudell E, Nambudiri G, Kraker AJ, Corey SJ. Therapeutic targeting of Src-kinase Lyn in myeloid leukemic cell growth. Leukemia. 1999; 13: 855-861. 18. Jordan CT, Upchurch D, Szilvassy SJ, et al. The interleukin-3 receptor alpha chain is a unique marker for human acute myelogenous leukemia stem cells. Leukemia. 2000; 10: 1777-1784. 19. Taussig DC, Pearce DJ, Simpson C, et al. Hematopoietic stem cells express multiple myeloid markers: implications for the origin and targeted therapy of acute myeloid leukemia. Blood. 2005; 106: 4086-4092. 20. Donella-Deana A, Cesaro L, Ruzzene M, Brunati AM, Marin O, Pinna LA. Spontaneous autophosphorylation of Lyn tyrosine kinase at both its activation segment and C-terminal tail confers altered substrate specificity. Biochemistry. 1998 Feb 3; 37(5): 1438-1446. 21. Robinson LJ, Xue J, Corey SJ. Src family tyrosine kinases are activated by Flt3 and are involved in the proliferative effects of leukemia-associated Flt3 mutations. 2005; Exp Hematol. 33: 469-479. 22. Contri A, Brunati AM, Trentin L, et al. Chronic lymphocytic leukemia B cells contain anomalous Lyn tyrosine kinase, a putative contribution to defective apoptosis. 2005; J Clin Invest. 115: 369-378. 23. Brdicka T, Pavlistova D, Leo A, et al. Phosphoprotein associated with glycosphingolipid-enriched microdomains (PAG), a novel ubiquitously expressed transmembrane adaptor protein, binds the protein tyrosine kinase csk and is involved in regulation of T cell activation. J Exp Med. 2000; 191: 1591-1604. 24. Ohtake, H., Ichikawa, N., Okada, M., Yamashita, T. Transmembrane phosphoprotein Cskbinding protein/phosphoprotein associated with glycosphingolipid-enriched microdomains as a negative feedback regulator of mast cell signaling through the FcepsilonRI. J Immunol. 2002; 168: 2087-2090. 25. Xu Y, Harder KW, Huntington ND, Hibbs ML, Tarlinton DM. Lyn tyrosine kinase: accentuating the positive and the negative. Immunity. 2005; 22: 9-18. 20

26. Tartaglia M, Martinelli S, Cazzaniga G, et al. Genetic evidence for lineage-related and differentiation stage-related contribution of somatic PTPN11 mutations to leukemogenesis in childhood acute leukemia. Blood. 2004; 104: 307-313. 27. Chen P, Levis M, Brown P, Kim KT, Allebach J, Small D. FLT3/ITD mutation signaling includes suppression of SHP-1. J Biol Chem. 2005; 280: 5361-5369. 28. Johan MF, Bowen DT, Frew ME, Goodeve AC, Reilly JT. Aberrant methylation of the negative regulators RASSFIA, SHP-1 and SOCS-1 in myelodysplastic syndromes and acute myeloid leukaemia. Br J Haematol. 2005; 129: 60-65. 29. Watanabe D, Ezoe S, Fujimoto M, et al. Suppressor of cytokine signalling-1 gene silencing in acute myeloid leukaemia and human haematopoietic cell lines. Br J Haematol. 2004; 126: 726-735. 30. Caligiuri MA, Briesewitz R, Yu J, et al. Novel c-CBL and CBL-b ubiquitin ligase mutations in human acute myeloid leukemia. Blood. 2007; 110: 1022-1024. 31. Sargin B, Choudhary C, Crosetto N, et al. Flt3-dependent transformation by inactivating c-Cbl mutations in AML. Blood. 2007; 110: 1004-12. 32. Harder KW, Parsons LM, Armes J, et al. Gain- and loss-of-function Lyn mutant mice define a critical inhibitory role for Lyn in the myeloid lineage. Immunity. 2001; 15: 603615. 33. Mermel CH, McLemore ML, Liu F, et al. Src-family kinases are important negative regulators of G-CSF dependent granulopoiesis. Blood. 2006; 108: 2562-2658. 34. Lannutti BJ, Minear J, Blake N, Drachman JG. Increased megakaryocytopoiesis in Lyndeficient mice. Oncogene. 2006; 25: 3316-3324. 35. Karur VG, Lowell CA, Besmer P, Agosti V, Wojchowski DM. Lyn kinase promotes erythroblast expansion and late-stage development. Blood. 2006; 108: 1524-1532. 36. Hu Y, Liu Y, Pelletier S, et al. Requirement of Src kinases Lyn, Hck and Fgr for BCRABL1-induced B-lymphoblastic leukemia but not chronic myeloid leukemia. Nat Genet. 2004; 36: 453-461. 37. Ptasznik A, Nakata Y, Kalota A, Emerson SG, Gewirtz AM. Short interfering RNA (siRNA) targeting the Lyn kinase induces apoptosis in primary, and drug-resistant, BCRABL1(+) leukemia cells. Nat Med. 2004; 10: 1187-1189. 38. Okamoto M, Hayakawa F, Miyata Y, et al. Lyn is an important component of the signal transduction pathway specific to FLT3/ITD and can be a therapeutic target in the treatment of AML with FLT3/ITD. Leukemia. 2007; 21: 403-410. 39. Martelli AM, Nyakem M, Tabellini G, et al. Phosphoinositide 3-kinase/Akt signaling 21

pathway and its therapeutical implications for human acute myeloid leukemia. Leukemia. 2006; 20: 911-928. 40. Zhu QS, Xia L, Mills GB, Lowell CA, Touw IP, Corey SJ. G-CSF induced reactive oxygen species involves Lyn-PI3-kinase-Akt and contributes to myeloid cell growth. Blood. 2006; 107: 1847-56. 41. Tamburini J, Chapuis N, Bardet V, et al. Rational for specific inhibition of both PI3K/AKT and mTORC1 activities in Acute Myelogenous Leukaemia. Blood. 2006; 108: 1904. 42. Billottet C, Grandage VL, Gale RE, et al. A selective inhibitor of the p110delta isoform of PI 3-kinase inhibits AML cell proliferation and survival and increases the cytotoxic effects of VP16. Oncogene. 2006; 25: 6648-59. 43. Leseux L, Hamdi SM, Al Saati T, et al. Syk-dependent mTOR activation in follicular lymphoma cells. Blood. 2006; 108: 4156-4162.

22

Legends to figures Figure 1: Analysis of the global protein tyrosine phosphorylation in normal HPCs and AML cells. A) The level of protein tyrosine phosphorylation in normal unstimulated and stimulated (GMCSF, SCF and IL-3 for 5 and 20 minutes or for 4 hours) CD34+ HPCs, fresh AML samples and AML cell lines was studied under similar conditions by Western Blot with the antiphosphotyrosine antibody 4G10. The membrane was stripped and reprobed for -actin as a loading control (lower panel) and for Lyn (black stars). Results shown are representative of experiments performed with normal CD34+ HPCs from 3 healthy donors and 38 AML samples (patients # 1; 2; 4-22; 24-28; 30; 31; 33-37, 56-60). B) A similar analysis was performed on fresh AML cells cultured alone or in the presence of PP2 (10 M) for 1 hour. Results shown are representative of experiments performed with 20 AML samples (patients # 1-20). Black stars indicate the position of Lyn.

Figure 2: Constitutive activating phosphorylation of SFKs in leukemic bulk and in immature CD34+CD38-CD123+ leukemic cells. A) Normal CD34+ HPCs unstimulated or stimulated with GM-CSF, SCF and IL-3, KG1 and 53 AML samples (patients # 1; 2; 4-22; 24-28; 30; 31; 33-37; 40-44; 47-53; 56-63) were analyzed by Western Blot using the phosphospecific antibody recognizing the activated form of SFK members (anti-pSrc Y416). The membranes were stripped and reprobed with the pan total Src antibody and a -actin antibody as loading control. Data shown are representative of 3 independent experiments. B) Flow cytometric analysis of a sample from patient # 66 using the four-color protocol as described in Material and methods. The left panel shows the CD34+ versus CD38 dot plot (CD34+CD38- blast cell population is represented in gate A). Middle panel shows the CD123+ and p-Src Y416 expression in gate A. The right panel shows a

23

sample of the leukemic bulk from patient # 66 analyzed both by flow cytometry and by Western Blot using the anti-pSrc Y416 antibody. C) Patient cells were cultured for 1 hour alone or in the presence of PP2 (10 M), then processed using four-colour immunostaining. The effect of PP2 on SFK phosphorylation was analyzed in the CD34+CD38-CD123+ blast sub-compartment (patients # 64-68) and in the leukemic bulk (patients # 20; 42-44; 47-50; 5253, 61-63). Results shown are representative of the different patients analyzed.

Figure 3: Characterization of SFK members in AML cells. A) The levels of Fyn, Lck, Hck, Fgr and Lyn from normal CD34+ HPCs and AML samples, were analyzed by Western Blot using specific antibodies (patients # 1; 2; 4-7; 9; 10; 12-16; 18; 19; 21; 22; 24-28; 30; 31; 33-41). The level of -actin was assessed as loading control (lower panel). Results shown are representative of the different patients analyzed. B) Protein lysates from KG1 cells and from AML samples (patients #18; 40; 43; 48) were immunoprecipitated using an anti-Lyn antibody. The immunoprecipitates (IP) were probed with anti-pSrc Y416. The membranes were then stripped and reprobed with a biotinylated anti-Lyn antibody revealed with streptavidin-HRP. Control immunoprecipitates were performed with protein A beads only. C) KG1, U937 and fresh AML samples (patients # 7; 15; 18; 45) were cultured after electroporation with Lyn siRNA or control siRNA. The Western Blot were performed 48-72 hours after electroporation. Results shown are representative of four experiments performed with KG1 cells and of the different patients analyzed. D) Immunolocalization of Lyn and active forms of SFKs was performed by confocal microscopy using specific antibodies in normal CD34+ HPCs and AML samples. Results shown are representative of 3 (for HPC, KG1 and U937) and 8 AML samples, patients # 8; 16, 21, 30, 40; 43; 47, 48) independent experiments. Original magnification, x 63, calibration bar = 5m.

24

Figure 4: Activity of PP2 on normal HPCs and AML cell proliferation and survival. A) Fresh AML cells (patients # 6; 51; 54) and normal HPC (right panel) were grown in clonogenic assays in the presence of increasing doses of PP2 (black bars, control; grey bars, 2 M; hatched bars, 5 M; white bars, 10 M). Results are presented as percentage of control and are mean SEM of duplicates. Results shown for normal CD34+ HPCs are mean SEM of two independent experiments performed in duplicate. IC50 values were 4.8, 3.5 and 7.7 M for patients # 6, 51 and 54, respectively. Five other patients (# 42, 47, 48, 50, 55) were also analyzed with IC50 values of 28.2, 9.5, 10.2, 14.5 and 0.5 M, respectively. B) AML samples (patients # 13, 42, 43, 44-46) were incubated with or without 10 M PP2 for 24 h, then processed for apoptosis studies using annexinV/PI staining and Western Blot analysis using anti-PARP and anti-caspase 3 antibodies. Results are presented as mean percentages ( SEM) of annexin V positive cells in treated versus untreated cells ( p<0.001). C) Fresh AML samples (patients # 40 and 48) were electroporated with Lyn siRNA or control siRNA and their clonogenic properties were analyzed. CFU-Ls numbers were scored at day 7. Results are expressed as absolute number of leukemic colonies and are mean SEM of duplicates.

Figure 5: Inhibition of SFKs affects the phosphorylation of mTOR targets in AML cells. A) KG1 cells were cultured for 1 hour alone (control) or in the presence of rapamycin (10 nM), AG1296 (25 M), AG490 (20 M) and PP2 (10 M). Cells were lysed and analyzed by Western Blot with the indicated antibodies. Results shown are representative of 3 independent experiments. B) Fresh AML cells (patients # 1-20) were cultured for 1 hour alone or in the presence of PP2 (10 M), lysed and analyzed as in A. Results shown are representative of the different patients analyzed. C) Fresh AML cells (patients # 7; 19; 21; 42; 43; 45; 49; 53; 59) were cultured alone or in the presence of PP2 (10 M) or LY294002 (25 M) for 4 hours, then processed for Western Blot analysis.

25

Figure 6: Silencing Lyn inhibits the mTOR pathway independently of Akt in AML cells. A) KG1 and U937 were cultured after electroporation with Lyn siRNA or control siRNA. After 72 hours, cells were processed for Western Blot analysis using appropriate antibodies. Results shown are representative of 4 independent experiments. B) Fresh AML samples (patients # 15, 18, 45) were cultured during 48 hours after electroporation with Lyn siRNA or control siRNA, lysed and analyzed by Western Blot using indicated antibodies.

Supplementary figures Figure 1: Analysis of the global protein tyrosine phosphorylation in AML subtypes The level of protein tyrosine phosphorylation in normal unstimulated CD34+ HPCs and in AML samples according to their cytogenetic or molecular background, was monitored by Western Blot with the anti-phosphotyrosine antibody 4G10. The membranes were stripped and reprobed with an anti--actin antibody as loading control.

Figure 2: SFK family members in AML; expression and activity. A) The levels of Fyn, Lck, Hck, Fgr and Lyn in normal CD34+ HPCs and AML samples, were analyzed by Western Blot using specific antibodies (patients # 7; 15; 19; 22; 25; 30-33; 3537). The level of -actin was analyzed as loading control (lower panel). Results shown are representative of the different patients analyzed. B) Lyn is a major tyrosine phosphorylated protein in AML. Lyn was depleted from KG1 and U937 cell lysates by immunoprecipitation. Lyn-depleted lysates (Lyn IP) were then analyzed by Western Blot with the antiphosphotyrosine 4G10 antibody. As a control total extract (TE) and the non-depleted lysates (Ctrl IP) were also analyzed. C) Lyn in vitro kinase assay. Lyn was immunopurified from 5x106 Peripheral Blood Lymphocytes (PBL), KG1 and AML patient cells as describe in Material and Methods, washed twice in lysis buffer and once with kinase buffer containing 50

26

mM Tris/HCl pH 7,4, 2,5 mM MnCl2 and 5 mM MgCl2). The beads were resuspended in 40 L of kinase buffer containing 10 Ci of [-32P] ATP (Perkin-Elmer) with cold ATP and

incubated at 37C for 30 minutes under shaking. Reactions were terminated by addition of 10 L of 5X Laemli buffer. The samples were submitted to a 8% SDS-PAGE gel and Lyn

autophosphorylation was evaluated by PhopshorImager analysis.

Figure 3: Effect of PP2 on KG1 cell cycle and proliferation A) KG1 cells were grown in clonogenic assays in the presence of increasing doses of PP2. Results are presented as percentage of CFU-L for each PP2 concentration relative to untreated cells and are mean SEM of three independent experiments performed in duplicate. B) KG1 cells were cultured in liquid medium alone or with increasing doses of PP2 for 4 days. Cell viability was quantified by MTT assay. Results are mean SEM of three independent experiments performed in triplicate. C) KG1 cells were cultured in liquid culture alone or with increasing doses of PP2 for 24 hours then processed for cell cycle analysis. Briefly, cells were washed in phosphate-buffered saline (PBS; NaCl 7.6 g/L; Na2HPO4 0.7 g/L; KH2PO4 0.2 g/L) with 5.5 mM glucose at 4C and fixed in 70% ethanol overnight at 4C. Cells were then resuspended in 1 mL PBS containing 50 g/mL propidium iodide and 100 U/mL RNAse A and incubated for 30 minutes at room temperature. The DNA content was monitored by flow cytometry (EPICS XL-MCL; Beckman Coulter, Villepinte, France). Percentages of cells in G0/G1, S, G2/M phases are indicated in each plot. Results shown are representative of three experiments.

27

Table 1: Characteristics of AML patients.

Patient n 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 WBCx109/L 46.5 1.1 1.8 216 7.6 162 148 142.5 225 28 54.5 125 33.3 167 98.6 107.7 91 70 20 18.9 57.8 35.7 3,3 21 23 11.2 11 18.9 1.3 33.5 211.8 13.3 10 42.1 39.6 98.6 142.5 3.1 106 4.4 88 50 188 39.5 38.5 195 14 41 60.4 136 13.3 82.7 1.4 10.6 19 FLT3 ITD 0 0 0 1 0 0 0 1 0 0 1 1 1 1 0 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1 1 0 1 1 1 0 nd nd 0 0 1 0 0 0 0 1 0 0 0 0 nd 0

Age,y 75 51 38 45 67 59 43 52 53 43 53 59 46 52 53 40 64 60 59 72 41 54 35 41 56 15 54 27 73 29 44 63 45 42 45 78 62 49 66 48 74 81 79 61 80 77 70 63 33 76 73 68 79 69 70

FAB 5 1 2 4 t-AML 4 4Eo 5 1 2 4 1 1 1 4Eo 5 4 2 4 2 4Eo 2 2 2 4Eo 2 2 3 3 2 2 3 3 3 t-AML 2 2 2 2 2 1 2 t-AML 2 2 1 1 t-AML 2 4 2 t-AML 2 2 t-AML

cytogenetic 46 XX 46 XX tetrasomy 8 46 XY 46 XY 46 XX 46 XY, inv (16) 46 XY, t(3;6) 46 XX 46 XY, t(8;21) 46XX Penta8; +(10); +(13) 46 XX 46 XX 46 XX, inv(16) 46 XY 46 XY, t(2;21) 46 XX 46 XX complex 46 XY, t(4;16) 47 XX, +(8) 46 XY, t(8;21) 46 XY, t(8;21) 46 XX, inv (16) 46 XY, t(8;21) 46 XY, t(8;21) 46 XX, t(15;17) 46 XY, t(15;17) 46 XX, t(8;21) 46 XY, t(8;21) 46 XX, t(15;17) 46 XY, t(15;17) 46 XY,t(15;17) 45 XX, inv(3), del(5) 46 XX, inv(3) 46 XX, t(3;6), 7qcomplex complex complex 46 XY, t(8;22),5q-,20q46 XY 47 XY, +(8) 46 XY 47 XY, +(8) 46 XX complex complex 46 X-Y+X , t(8;21) 47 XX, +(11) 46 XX 45 XY, -(7) complex 46 XY 46 XY, (7)q-

D835 0 0 0 0 0 nd 0 0 nd 0 0 nd 0 0 0 0 nd 0 0 nd 0 nd 0 0 0 nd 0 nd nd 0 0 1 nd 0 0 0 0 nd nd nd nd 1 0 0 nd 0 nd 0 0 1 nd nd 0 nd nd

28

56 57 58 59 60 61 62 63 64 65 66 67 68

56 53 19 66 60 76 68 82 56 64 64 44 79

1 1 5 2 2 1 2 2 1 4 1 4 1

26,9 27,9 15,2 5,8 41 6,5 1,2 1,6 160 20 120 120 1,2

nd 46 XY 5q-, 17+, -20, t(5;17) complex 46 XY 46 XX complex ;5q- ;17p-7;+8 complex complex 46 XY 46 XX complex

1 0 0 0 0 nd 0 nd 0 0 0 1 nd

nd 0 nd nd nd nd nd nd nd nd nd nd nd

FAB indicates French-American-British classification; WBC, white blood cell count; FLT3-ITD, FLT3-internal tandem duplication; FLT3-D835, FLT3 mutation in the activation loop; Diag, diagnosis; t-AML, transformedAML; ND, not done.

29

Figure 1 A
CD34+
M W (kDa) : T0 5 20' 4h 181.8 115.5 82.2 64.2 48.8 37.1 #1 #2 #4 #5 #7 #16 #15 #17

AML Patients
M W (kDa) : 250 150 100 75

KG1 U93 7 UT7 -epo

* *

* *

50 37 25 -Actin

B
#1
PP2
M W (kDa) :

#4
PP2

#5
PP2

#18
PP2
* *

181.8 115.5 82.2 64.2 48.8 37.1 -Actin

-Actin

Dos Santos et al

Figure 2 A
T0

CD34+
5 20 4h

KG1
#1

AML Patients
#4 #5 #16 #15 #17

* *

p-Src (Y416)

Pan total Src

-Actin

AML CD34+ CD38- CD123+ #66 A

A=CD34+CD38RFI=6.8

AML BULK #66

RFI=6.4

p-Src (Y416)

C
AML CD34+ CD38- CD123+ #66
+ PP2 control - PP2 + PP2 control

AML BULK #66

- PP2

Dos Santos et al

Figure 3 A
CD34+
#1 #2

AML Patients
#4 #5 #16 #21 p59 Fyn

p56 Lck p61 Hck p59 Hck p55 Fgr p56 Lyn p53 Lyn -Actin

B
KG1
Lyn IP Ctrl IP

#18
Lyn IP Ctrl IP

#40
Lyn IP Ctrl IP

#43
Lyn IP Ctrl IP

Ctrl IP

Lyn IP

KG1
Lyn s iR NA RNA

U937
Lyn s iR NA RNA

#7
Lyn s iR NA RNA

#18
Lyn s iR NA RNA

Ctrl si

Ctrl si

Ctrl si

Ctrl si

#48

p-Src (Y416)

Lyn

Lyn

p-Src (Y416)

-Actin

Dos Santos et al

Figure 3 D
Lyn p-Src(Y416) Merge Profile

Normal CD34+
5 m

#47

U937

KG1

Lyn/p-Src

Dos Santos et al

Figure 4 A
120 CFU-L number (% of control recovery) CFU-L number (% of control recovery) 100 80 60 40 20 0 #6 #51 #54 #43 Annexin V positive cells (%) 50 40 30 20 10
Ctrl + PP2 Caspase 3 full length

100 80 60 40 20 0 CFU-GM BFU-E CFU-M

PP2

60

***

PARP Cleaved PARP

AML samples (n=6)

C
CFU-L (absolute number) 2000 1500 1000 500 0

CFU-L (absolute number)

#40

#48

3000 2000 1000

Lyn siRN A

Lyn siRN A

0
Ctrl siRN A

Ctrl siRN A

-Actin

Dos Santos et al

Figure 5 A
Rap amy cin Con trol

KG1
AG1 296

AG4 90

PP2

p85S6K p70S6K

P-p70S6K (Thr 389)

Total p70S6K

P-4E-BP1 (Thr 70)

Total 4E-BP1

-Actin

B
AML cells #1
PP2 C

#2
PP2 C

#4
PP2 C

PP2

#5

P-p85S6K P-p70S6K Total p70S6K

P-p70S6K (Thr 389)

P-4E-BP1 (Thr 70)

Total 4E-BP1

-Actin

Dos Santos et al

Figure 5 C
AML cells #19
LY2 940 02

#49
LY2 940 02

PP2

PP2

p-Src (Y416)

P-Akt (Ser 473)

Total Akt

-Actin

Dos Santos et al

Figure 6
Lyn s iR NA Lyn s iR NA
Lyn P-4E-BP1 (Thr 70) Total 4E-BP1 P-RPS6 (Ser 235/236) P-Akt (Thr 308) -Actin

KG1
RNA

U937
RNA Ctrl si

Ctrl si

#15
Lyn s iR NA RNA

#18
Lyn s iR NA RNA

#45
RNA Lyn s iR NA
Lyn P-4E-BP1 (Thr 70) Total 4E-BP1 P-RPS6 (Ser 235/236) Total RPS6 P-Akt (Ser 473) Total Akt P-Erk 1/2 -Actin

Ctrl si

Ctrl si

Ctrl si

Dos Santos et al

ABSTRACT Telomerase regulation in normal and leukemic haematopoietic cells. Human telomerase is a ribonucleoprotein DNA polymerase comprising a catalytic protein subunit, telomerase reverse transcriptase (hTERT), which represents the rate-limiting state in telomerase activity, and a RNA template (hTR). The primary defined function of telomerase is to elongate telomere at the end of chromosomes and allow cells to bypass replicative senescence. Recently, other important cellular functions have been attributed to telomerase, including cell proliferation, genetic stability, protection against apoptosis and cell differentiation. hTERT is highly expressed in cancer cells including acute myeloid leukaemia (AML), and in proliferative tissues such as haematopoietic cells. Previous studies have indicated that telomerase activity is low in primitive haematopoietic cells, but increases upon stimulation with a mixture of cytokines in parallel to cell expansion, and then declines progressively during differentiation. These observations suggest a function for telomerase in haematopoiesis. The aim of our study was to assess hTERT regulation by HGF in normal and leukemic cells. In the first part of the study, we showed that in AML cells, treatment with TNF- induces a decrease in hTERT gene transcription through a ceramide/JNK pathway, and that coincubation with GM-CSF can inhibit the effect of TNF-. Interestingly, in normal haematopoietic progenitors, TNF- also down-regulate hTERT gene expression alongside with a decrease in proliferation and an increase in differentiation. In the second part of the study, we investigated whether hTERT can be regulated during erythropoiesis, by erythropoietin (EPO) and TGF-, wich are respectively the major positive and negative regulators of erythropoiesis. As a result, we demonstrated that EPO can stimulate hTERT transcription through a JAK2/STAT5/c-myc pathway, and that TGF- counteracts this effect through Smad3 activation. Moreover, hTERT inhibition by ectopic expression of a dominant negative, reveals that EPO-mediated hTERT regulation serves neither for the proliferative response to EPO, nor to enhance cell survival, but can play a role in long term erythroid expansion. In conclusion, the compiled results produced clearly suggest that telomerase can be regulated by haematopoietic cytokines, and that all events leading to inhibition of hTERT expression may potentially alter haematopoiesis and lead to medullar insufficiency found in myelodysplastic syndromes for instance.

Key words : telomerase, haematopoiesis, cytokines, growth factors, leukaemia.

AUTEUR : Nas PRADE-HOUDELLIER TITRE : Rgulation de la tlomrase dans les cellules hmatopotiques normales et leucmiques. DIRECTEUR DE THESE : Docteur Vronique DE MAS RESUME : La tlomrase est une transcriptase inverse constitue dune sous-unit catalytique hTERT, facteur limitant de lactivit tlomrase, et dune sous-unit ribonuclique hTR. Sa fonction primordiale est de compenser lrosion des tlomres lors des divisions cellulaires. Rcemment, dautres rles de la tlomrase ont t dmontrs notamment dans la protection contre lapoptose (Gorbunova et al, 2002) ou dans la diffrenciation (Armstrong et al, 2005). Cette enzyme est fortement exprime dans la plupart des cellules tumorales dont les leucmies aigus mylodes (LAM) ainsi que dans les cellules renouvellement rapide comme les cellules de lhmatopose. Ainsi, lactivit hTERT, faible dans les cellules souches hmatopotiques, augmente lorsquelles sont mises en prsence dun mlange de cytokines capables dassurer leur expansion puis diminue en fin de maturation (Engelhardt et al, 1997). Sur la base de ces considrations, il est logique de supposer que hTERT est critique pour assurer lexpansion de lhmatopose. Lobjectif de notre travail tait dtudier les voies de rgulation de hTERT par les facteurs de croissance et les cytokines inhibitrices de lhmatopose dans les cellules normales et leucmiques. Dans la premire partie de notre tude, nous avons montr que le TNF-, cytokine inflammatoire, inhibe dans des cellules de LAM, la transcription du gne hTERT en activant une voie cramide/JNK. Cette rgulation ngative est corrle un processus de snescence et une instabilit chromosomique (Beyne-Rauzy, 2004). Le GM-CSF permet de bloquer cette inhibition. De plus, le TNF- rgule ngativement hTERT dans les progniteurs hmatopotiques normaux avec une diminution de la prolifration et une acclration de la diffrenciation. Dans la deuxime partie de notre tude, nous nous sommes plus particulirement intresss la rgulation de hTERT au cours de lrythropose par lrythropotine (EPO) et le TGF-, facteurs activateur et inhibiteur respectivement. Dans ce modle, lEPO active transcriptionnellement hTERT via la voie JAK2/STAT5/c-myc. Le TGF- active Smad3 et bloque cette activation. De plus, linhibition de hTERT par expression dun dominant ngatif a rvl que cette enzyme ne joue pas de rle direct dans la prolifration induite par lEPO ou la protection contre lapoptose, mais permet lexpansion cellulaire long terme. Lensemble de nos rsultats suggre que la tlomrase peut tre rgule par les cytokines de lhmatopose, et que tout vnement conduisant la rpression de lexpression de cette enzyme devrait induire de profondes altrations de lhmatopose, pouvant avoir des consquences fonctionnelles importantes telles que linstabilit chromosomique, ou linsuffisance mdullaire retrouve dans certaines pathologies comme les mylodysplasies.

MOTS-CLES : Tlomrase, hmatopose, cytokines, facteurs de croissance, leucmie. DISCIPLINE : Cancrologie LABORATOIRE : INSERM U563 Centre de Physiopathologie Toulouse Purpan