en fr

HAL Id: tel-01557533

https://tel.archives-ouvertes.fr/tel-01557533

Submitted on 6 Jul 2017

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

and a long non-coding RNA in plasmacytoïd dendritic

cell acute leukaemia

Neda Hoghoughi

To cite this version:

Neda Hoghoughi. Functional characterisation of a novel t(3;5) translocation targeting the Gluco-corticoïd receptor gene and a long non-coding RNA in plasmacytoïd dendritic cell acute leukaemia. Agricultural sciences. Université de Grenoble, 2014. English. �NNT : 2014GRENV073�. �tel-01557533�

THÈSE

Pour obtenir legrade de

DOCTEUR

DE

L

’UN

IVERS

ITÉ DE

GRENOBLE

Arrêté ministériel : 7août 2006

Présentée par

Neda

HOGHOUGH

I

Thèse dirigée par Mary CALLANAN

préparée au sein de L'équipe 7 «Voies Oncogéniques des Hémopathies Malignes »

Centre de Recherche INSERMU823 –Université Joseph Fourier Grenoble 1 –InstitutAlbert Bonniot

dans l'ÉcoleDoctorale Chimie et Science du Vivant

Caractér

isa

t

ion

fonct

ionne

l

le

d

'une

nouve

l

le

trans

locat

ion t(3

;5)(q21

;q31

)

,

c

ib

lant

le

gène

du

récepteur

aux

g

lucocort

ico

ïdes

e

t

un

ARN

non-codant

,

dans

la

leucém

ie

a

iguë à

ce

l

lu

les

p

lasmocy

to

ïdes

dendr

it

iques

Thèse soutenue publiquement le19 décembre 2014 devant lejurycomposé de :

Pr. Philippe GAULARD

PU-PH, InsermU955 -Université Paris-Est, Marne-la-Vallée, Rapporteur Dr. Marina LAFAGE-POCHITALOFF

MCU-PH, CHU de Marseille -Université Aix-Marseille, Rapporteur Pr. Dominique LEROUX

PU-PH,Inserm U823 -CHU de Grenoble -UJF ,Président Dr. Mary CALLANAN

MCU-PH, Inserm U823-UJF-CHU de Grenoble, Examinatrice Invités:

Dr. Christine LEFEBVRE PH, CHU de Grenoble –UJF Dr. Anouk EMADALI Chercheur, CHU de Grenoble

« L’expérience très enrichissante que j’ai pu vivre au cours de ces quatre années m’a prouvé que la thèse est loin d'être un travail solitaire. En effet, je n'aurais jamais pu réalisercetravail doctoralsanslesoutien d'un grand nombre de personnes dontla générosité,la bonne humeuretl'intérêt manifestéàl'égard de marecherche m'ont permis de progresser dans cette phase exaltante mais néanmoins délicate ».

En premier lieu, je tiens à remercier ma directrice, la Dr. Mary Callanan, pour la confiance qu’elle m'a accordée en acceptant d'encadrer cettethèse, pour ses multiples conseils, et pourtoutesles heures qu’elle a consacrées à diriger cette recherche malgré uneimportante charge detravail. Sa compétence etsarigueurscientifique m’ont beaucoup appris,ils ont été et resteront des moteurs dans montravail de chercheur.

Je souhaiterais exprimer ma gratitude envers les membres de mon jury de thèse, la Pr Dominique Leroux,le Pr Philippe Gaulardetla Dr Marina Lafage-Potchitaloff pour avoir accepté d’évaluer mestravaux dethèse. Jeremercie égalementla Dr Christine Lefevbre etla Dr Anouk Emadali pour m’avoir fait part deleur expertise sur cetravail de recherche.

Je remercie le Ministère de l’Enseignement Supérieur et de la Recherche et l’école doctorale del’université de Grenoble I, pour m’avoir accordé une allocation de rechercheallouéelors destrois premièresannées de mathèse,ainsi quelasociété française d’hématologie quiafinancéen partie ma dernièreannée dethèseet m’a donnél’opportunité de présenter mestravauxaucongrèsannuel dela SFHen mars 2014. Je tiens également à remercier chaleureusement l’association Aramis pour son soutien.

Jetiens àremercierla Pr. Dominique Leroux pour avoir effectuéles bases de recherches nécessaires et primordiales permettant la réalisation de ce projet. Je remercie égalementle Dr Rémi Gressin pour ses conseils et son encouragement.

Un mercitout particulierà Samuel Duley pourtouslestravaux qu’ilaeffectués auparavant danslecadre dece projetet pour m’avoiraccompagnéavecsa bonne humeur dèsle premierjour de monarrivé.Jeremercieégalement Anouk Emadali pourtoutesa participationàcetravail derechercheet biensûr poursesconseils

précieux.

Jetiens à remercierle Dr Saadi Khochbin pourtous nos échanges enrichissants, son encouragement pendantles moments difficilesetsonsoutien moraletscientifique pendanttoutes ces années.

remercie pour m’avoirsupporté dansson bureau pendantles dernières mois dela rédaction de mathèse, et pour avoir partagé avec moi ses expériences et ses conseils.

Je remercie chaleureusement Sième Hamaidia pour toute l’aide qu’elle m’a apportée, poursonsourireetsa bonne humeur, Martine Chauvetet Patricia Betton pourleur patience,leur gentillesse etleur rigueur.

Merci à Sarah, Azadeh, Estelle, Amandine et Aurélie pour tous les bons souvenirs queje garde decettethèse grâceàleur présence.Jeremercieégalementtousles membres d’équipe au CHU pourleur disponibilité,leur souplesse seleur support.

Jetiens àremercier égalementla Pr Claire Vourc’h, Virginie, Solenne, Lydia, Gaetan, Jessica et Catherine pour m’avoir épaulé en début de thèse avec les hauts et les bas durant cette période et aussi pour les bons moments passés à leurs côtés, les repas,les gâteaux, merci pourtous ces souvenirs que vous melaissez.

Ces remerciements seraientincomplets sije n'en adressais pas àl'ensemble des membres del’institut pourleur soutienlogistique et moral ainsi que pourlatrès bonne ambiance quej'aitrouvée àl’institut.

Ma profondereconnaissance va à ceux qui ont plus particulièrement assuré un soutienaffectiflorscetravail doctoral mes parents, masœuret ma grand-mère qui m’onttoujours épaulé dans ce projet. Je sais que mon absence a étélongue etj’espère pouvoir unjourrattraperleretardaccumulé. Flavien, merci de m’avoirsoutenuet encouragétout aulong de ces années.

Abs

trac

t

Blastic plasmacytoid dendritic cell neoplasm(BPDCN) is an incurable malig-

nancy for which disease mechanisms areunknown. Here,we identifytheNR3C1 gene

(5q31), encodingthe glucocorticoidreceptor(GCR), and along,intergenic, non-coding RNA gene (named herelincRNA-3q), respectively, as targets for genetic alteration or transcriptional deregulation in BPDCN. NR3C1 translocation/deletion was associated tocriticallyshort survival in BPDCNand toabnormal activity of GCR, EZH2,and

FOXP3 generegulatory networks. LincRNA-3q, was foundto encode anuclear,

non-coding RNA thatis ectopicallyactivated in BPDCN and high-riskAML. Depletion of lincRNA-3q inmyeloid cancercells inducedcell cycle arrest, coincidentto suppression of E2F1/Rb and leukemiastem cell-specific gene expression signatures. BET bromod-omain proteininhibition could selectively suppresslincRNA-3q indicating atreatment strategy for counteracting oncogenic activity ofthis non-coding RNA. Thus,this work defines a new framework for understanding disease pathogenesis and treatment res

Contents

Abbreviations xiii

1 Dendritic cells 1

1.1 Hematopoiesis- General principles... 1

1.2 Dendritic cell lineage... 2

1.3 Dendritic Cell development ...3

1.4 Dendritic Cell Subsets... 4

1.4.1 Plasmacytoid Dendritic Cells... 4

1.4.2 Conventional Dendritic Cells... 6

1.4.3 InflammatoryMonocytes ...6

1.4.4 Langerhans Cells... 7

1.5 NK lineage... 7

1.6 Transcriptional control ofPlasmacytoid Dendritic Celldevelopment ...8

1.6.1 Ikaros... 8

1.6.2 E2-2... 10

1.6.3 SPI-B... 10

1.6.4 Interferonregulatory factor 8... 11

1.6.5 RUNX2 ...11

1.6.6 L-Myc... 12

2 Blastic plasmacytoid dendritic cell neoplasm 15 2.1 Classification and clinicalfeatures... 15

2.2 Molecular pathogenesis of BPDCN... 17

2.2.1 Chromosomal aberrations... 17

2.2.2 Recurrent gene mutations in BPDCN... 19

2.2.2.1 Mutations oftumor suppressor genes... 19

TP53... 20

RB protein families ...20

2.2.2.2 Gene mutations affecting cell signaling... 22 v

FLT3 ...22

BRAF... . . . 23

2.2.2.3 Mutations intranscriptionfactors . . . 24

Ikaros... . . . 24

2.2.2.4 Mutations in epigeneticregulators . . . 25

Epigenetics: an emerging hallmark of cancer... 25

General overview on epigenetics... 26

Epigeneticreaders, writers and erasers... 28

2.2.2.5 Mutations affecting DNMT3A,TET2 andIDH2... 29

DNMT3A... 29

TET2 ... 32

IDH2... 34

2.2.2.6 Mutations affecting histone-modulatingfactors ...36

MLL ... 36

Mutations affecting the Polycombrepressive complex PRC2... 37

PRC2 catalytic unit: EZH2 ... 38

PRC2 accessory sub-unit: ASXL1 ... 40

2.2.2.7 Emerging epigenetic players: non-codingRNAs ... 42

Long non-coding RNA ... 43

Mechanisms offunction of nuclearlong non-coding RNA ... 44

Physiological roles oflncRNAexpression in generegulation ...45

X-chromosome inactivation...45

Genomicimprinting... 47

Regulation ofHOX genes... 48

Long noncoding RNA in cancerdevelopment ...48

2.2.3 Gene expression profiling of BPDCN -insightsto molecular pa tho-genesis... 50

3 BPDCN clinical management 53 3.1 Conventional AML typetherapy...54

3.2 Conventional ALL typetherapy...55

3.3 Conventional Lymphoma typetherapy...55

Glucocorticoidsreceptor signaling... 56

Nucleocytoplasmic translocationof GCR... 58

3.4 Emerging therapeuticstrategies ...59

3.4.1 Inhibitors of signalingpathways ...59

3.4.1.2 NF-kBpathway ...60

3.4.2 Epidrugs... 61

3.4.2.1 Hypomethylating agent 5-Azacytidine ...61

3.4.2.2 BET bromodomaininhibitors... 63

4 Scientific arguments 65

5 Results 69

6 General discussion and perspectives 121

          Bibliography 133

L

ist

of

F

igures

1 Hierarchical model of hematopoiesis inthe adult bonemarrow ...2

2 Organization ofthe dendriticcell (DC)network ...4

3 Growth factors andtranscriptionfactors regulating DC differentiation . 9

4 The central role ofthe retinoblastoma protein (pRb) incell-cycle pro-gression... 21

5 A schematic diagram ofthe FLT3 receptor... 23

6 The hallmarks of cancer... 26

7 Overall structure ofthe epigenomein humancells ... 27

8 Enzymes involvedinDNA and histone modificationpathways ...30

9 Associated mutations inthe DNA methylation anddemethylation pathway 35 10 Schematicrepresentation ofthe PRC2 holoenzyme at chromatin... 38

11 Modelsfor nuclear lncRNAfunction... 46

12 Structure ofthe glucocorticoidreceptor... 58

13 Proposed mechanism of action of Azanucleosides... 62

14 The expression andactivity ofmammalian E2Ffamily members during thecell cycle ...125

15 A proposed model ofthefunctional consequence of lincRNA-3q expres-sion inleukemiccells ... 127

16 Schematic diagram of RAP... 128

17 Schematic diagram ofCHART ... 129

18 Schematicrepresentation of dChIRPtechnique... 130                         ix

L

ist

of T

ab

les

1 Therole of mutations in epigeneticregulators in myeloid malignancies. 42 2 Regulatory ncRNAs grouped bysize ... 43                                                                                      

Abbrev

iat

ions

2HG 2-Hydroxy Glutarate

5-hmC 5-hydroxy methyl-Cytosine

5-mC 5-methyl-Cytosine

AF Activation Function

ANRIL Antisense Non-coding RNA intheINK4 Locus

Airn Antisense insulin-likegrowth factor 2receptorRNA

AITL AngioImmunoblastic T-cell Lymphoma

ALL Acute Lymphoblastic Leukemia

Allo-SCT Allogeneic StemCell Transplantation

AML Acute Myeloid Leukemia

APC Antigen-PresentingCells

APL Acute Promyelocytic Leukemia

Ara-C Cytosine arabinoside

Asx Additional sex combs

ASXH amino-terminal ASX Homology

ASXL1 Additional sex combs-Like1

Auto-SCT AutologousStemCell Transplantation

BD BromoDomain

BDCA2 BloodDendritic Cell Antigen 2

BER Base-Excision Repair

BET Bromodomain andExtraTerminal domain

bHLH basic Helix-Loop-Helix

BM Bone Marrow

BPDCN BlasticPlasmacytoidDendritic Cell Neoplasm

BST2 Bone Marrow StromalAtntigen2

CBX 7 ChromoBoX 7

cDC conventional Dendritic Cell

CDP Common Dendritic Cells Progenitor

CDR common Deleted Region

CEBPA CCAAT/Enhancer-BindingProteinAlpha

CGH Comparative Genomic Hybridization

CHART Capture Hybridization Analysis of RNA Target

ChIP-seq ChromatinImmunoPrecipitationexperiments followed bysequencing

ChIRP ChromatinIsolation by RNA Purification

CLP Common Lymphoid Progenitor

CMML Chronic Myelomonocytic Leukemia

CMPs Common Myeloid Progenitor

CN-AML Cytogenetically Normal-Acute Myeloid Leukemia

CNS Central Nervous System

DBD DNA-Binding Domain

DC Dendritic Cell

dChIRP domain-specific ChromatinIsolation byRNA Purification

DHS DNaseI-HyperSensitive

DLBCL DiffuseLarge B-Cell Lymphomas

DNMT DNA MethylTransferase

DT DiptheriaToxin

EHMT2 Euchromatic Histone-lysine N-MethylTransferase 2

ENL ElevenNineteenLeukemia

ERK Extracellular signal-Regulated Kinase

ETP Enhancer of Trithorax andPolycomb

EVI1 Ecotropic Viraliintegrationsite 1

EZH2 Enhancer of Zeste Homologue 2

FAIRE-seq Formaldehyde-Assisted Isolation ofRegulatory Elements-sequencing

FAL1 Focally Amplified LncRNA on chromosome1

FISH FluorescenceIn SituHybridization

FLT3 Fms-LikeTyrosine kinase 3

FOXP3 Forkhead box P3

GATA1 Erythroidtranscription factor orGATA-binding factor 1

GC GlucoCorticoid

GCR GlucoCorticoid Receptor

GCRFP GlucoCorticoid Receptor FfusionProtein

GEP Gene ExpressionProfile

Gfi-1 Growthfactorindependent 1

GM-CSF Granulocyte-Macrophage Colony-Stimulating Factor

GMP Granulocyte/MacrophageProgenitor

GRE Glucocorticoid-Response Element

GSEA Gene Set Enrichment Analysis

GSK-3 GlycogenSynthase Kinase3

HAT HistoneAcetylTransferase

HDAC HistoneDeACetylase

HDM HistoneDeMethylase

HMT HistoneMethylTransferase

HOX HomeOboX

HSC Hematopoietic StemCell

HSP Heat Shock Proteins

HSPC Hematopoietic Stem andProgenitorCell

ICL Interstrand Cross-Link

ICR Imprinting Control Region

IDH IsocitrateDeHydrogenase

IFN1 InterFeroN,type1

Igf2r Insulin-likeGrowth Factor 2 Receptor

IHC ImmunoHistoChemistry

IL InterLeukin

IRF 8 Interferon Regulatory Factor 8

ITD Internal Tandem Duplication

IκB Inhibitor of NF-κB

JAK2 JAnus Kinase2

JM JuxtaMembrane domain

K Tyrosine-Kinase

Kcnq1ot1 Kcnq1ot1 overlapping transcript1

KI Tyrosine-KinaseInsert

KIT Tyrosine-protein kinaseKIT or CD117

KRAS KirstenRAtSarcoma viral oncogene homolog

LBD Ligand-Binding Domain

LC Langerhans Cell

LILRA4 Leukocyte Immunoglobulin-LikeReceptor subfamily A member 4

lincRNA Long intervening/intergenic noncoding RNA

lncRNA Long noncoding RNA

LT-HSC Long-Term-Hematopoietic StemCell

M-CSFR Macrophage Colony-StimulatingFactorReceptor

MDP Macrophage and Dendritic-cell Progenitor

MDS MyeloDysplastic Syndrome

MDS/MPN MyeloDysplastic Syndrome/MyeloProliferative Neoplasm

MEP Megakaryocyte/Erythroid Progenitor

MHC Major HistocompatibilityComplex

MLL MixedLineageLeukemia gene

MNase MicrococcalNuclease

MPL MyeloProliferative Leukemia

MPP MultiPotential Progenitor

ncRNA noncoding RNA

NDR Nucleosome-DepletedRegion

NF-κB Nuclear Factor-κ  B

NGS New Generation Sequencing

NK Natural Killer

NPM1 NucleoPhosMin1 gene

NRAS Neuroblastoma RAS viral(v-ras) oncogene homolog

O-GlcNAc O-linkedβ  -D-N-AcetylGlucosamine

OGT O-GlcNAcTransferase

ORF Open Reading Frame

p-ALL pediatric Acute Lymphoblastic Leukemia

PB Peripheral Blood

PBMC Peripheral BloodMononuclear Cell

PCAT ProstateCancer–AssociatedncRNA Transcript

PcG Polycomb Group

PCR Polymerase Chain Reaction

pDC plasmacytoidDendritic Cell

PHD PlantHomeoDomain

PTCL Peripheral T-Cell Lymphoma

PNAcqPCR Peptide Nucleic Acid clampreal-timePolymerase Chain Reaction

Pol II Polymerase II

PRC2 Polycomb Repressive Complex 2

Pre-DC Precursor Dendritic Cell

PTD Partial Tandem Duplications

PTM Post-Translational Modification

PTPN1 ProteinTyrosine Phosphatase, Non-receptortype 1

RA Retinoic Acid

RAP RNA Antisense Purification

RAR Retinoic Acid Receptor

RB RetinoBlastoma

RBD Ras BindingDomain

RBL1 RetinoBlastoma-Like1

RBP RNA BindingProtein

RIP RNA Immunoprecipitation Assay

RNA-seq RNA-sequencing

SET Su(var)3–9/Enhancer ofzeste/Trithorax

shRNA short hairpin RNA

SIGLEC-H Sialicacid-bindingImmunoGlobulin-likeLECtinH

SLO Secondary Lymphoid Organs

SNP SingleNucleotide Polymorphism

T-ALL Tcell Acute Lymphoblastic Leukemia

TBP TATA-BindingProtein

TET Ten-ElevenTranslocation

TET2 Tet methylcytosine dioxygenase2

TF Transcription Factor

TKD Tyrosine KinaseDomain

TLR Toll-LikeReceptor

TM TransMembrane domain

Topo II Topoisomerase II

TrxG Trithorax Group

TSS Transcription StartSite

WES Whole-Exome Sequencing

WHO World-Health Organization

WT1 Wilms-Tumor Proteintextbf1

XCI X Chromosome Inactivation

Xi X-inactive

Xic X-inactivation centre

Xist X-inactive specific transcript

YY1 YinYong 1

Chapter 1

Dendr

it

ic

ce

l

ls

1

.1 Hematopo

ies

is -

Genera

l

pr

inc

ip

les

Blood cells are divided into three lineages erythroid (erythrocytes), lymphoid (B and T cells), and myeloid (granulocytes, megakaryocytes, and macrophages). Thesecells are produced and constantly regenerated inthe bonemarrow from hematopoietic stem

cells (HSCs) through a series of progenitor stages (by a process called

hematopoiesis)(Figure 1). The current “deterministic model” for hematopoiesis proposes that Long- Term-Hematopoietic Stem Cells (LT-HSCs) which display a unique ability for life-long self-renewal give rise to Short-Term-Hematopoietic Stem Cells (ST-HSCs), which produce multipotential progenitors (MPPs) that have lost all self-renewal potentialbut are still able to generate all hematopoieticlineages. From these derive thecommon lymphoid progenitors (CLPs), whichgive riseto T, B, NK and certain dendritic cell (DC) cell subsets, and common myeloid progenitors

(CMPs)(which give rise to granulocyte/macrophage progenitors (GMPs)),

megakaryocyte/erythroid progenitors(MEPs),mast cell and basophil progenitors. This

process istightlycontrolled by lineage and differentiation stage specific

transcription factors that modulate gene expression programs to

permanently generatethecorrect number ofspecific cell typesata given time and place.

FIGURE 1: Hierarchical model of hematopoiesis in the adult bone marrow. From

(Wanget Wagers, Nat Rev Mol Cell Biol, 2011)  

A comprehensive analysis ofthe molecularcontrolandthe precise interplay of these cell fate decisions are crucial for a full understanding of normal and malignant hematopoiesis (Rieger and Schroeder, 2012). Disruptions of thesecell fate decisions underlie hematological disordersincluding blood cancers.

1

.2 Dendr

it

ic

ce

l

l

l

ineage

Dendritic cells (DCs) were first describedin 1973 by RalphM. Steinman andZanvil A. Cohn (Steinman and Cohn, 1973). DCs are antigen-presentingcells (APCs) ofthe mammalian immune system andthey play acritical roleintheregulation ofthe adap-tiveimmuneresponse. These cells arerecognized as the mostimportant APCs with the abilityto migrateinlymphatic system andto prime naive Tcells in secondary lym-phoid organs (SLOs) (Banchereau and Steinman, 1998).Therefore, they constitute a front-line defense against invading pathogens. DCsarelocatedthroughoutthe body and form a sophisticated and complex network thatallows themto communicatewith

different populations oflymphocytes,therebyformingan interface betweentheexternal environment andtheadaptive immune system(Belz and Nutt, 2012).

1

.3 Dendr

it

ic

Ce

l

l

deve

lopment

Inthe bone marrow (BM) a fraction of CMPexpresses FMS-relatedtyrosine kinase 3 (FLT3) anddifferentiates into more-restricted macrophage andDC progenitors(MDPs).

MDPs are thought to be the direct precursorto commonDC progenitors (CDPs) and

give risetotheDC lineages. CDPs produce precursor DCs(pre-DCs) andplasmacytoid DCs(pDCs), whichthen exit the bonemarrow andtravelthroughthe blood and migrate toward the secondary lymphoidorgans andnon-hematopoietic tissues. Alimited pro-portion of DCs may also derive from CLPsinthe bonemarrow and from early Tcell progenitorsinthe thymus(Belz and Nutt, 2012). The only subsets foundinthe spleen are lymphoidtissue-resident DCs, which understeady-stateconditionsarisefrom pre-DCs. This population is composed ofthree conventional DC subsets, namely cluster of differentiation 4+ _(CD4+_{) DCs, CD8α}+ _{DCsand CD8α}−_CD4− _doub

le-negative (DN) DCs (Figure 2).

Peripherallymph nodes contain CD8α+ _{and CD8α}− _{DC populationsbut arealso}

populated by two groups of migratory DCs;the dermalDC population whichisbroadly

composed of CD11b+_{and CD103}+ _{DCs, and, migrate through the lymphatic to the}

lymph node and Langerhanscells thatdevelop inthe epidermis and migratethrough the basement membrane tothelymph nodes viaterminallymphaticvessels thatarise in the dermis. Monocytes arrive at tissues from the blood and in response to inflammation theycan develop intomonocyte-derived DCs, which adoptmany ofthe characteristics of conventional DCs(Figure 2).

In ordertoprovide theirprotective function, specialized DCs subsetshave evolved andthis segregation wasinitially based ontheir distinct patterns of cell-surface molecule

FIGURE 2: Organization of the dendritic cell (DC) network and key surface pheno-

typemarkers of different DC subsets, delineated on thebasis of theirlocalization in secondary lymphoidtissues.From (Belz and Nutt, Nat Rev, Immunol, 2012)

Different subsets of DCs are composed of: plasmacytoid DCs (pDCs),

conven-tional DCs (cDCs), monocyte-derived inflammatory DCs and Langerhanscells (LCs)

(Belz and Nutt, 2012).      

1

.4 Dendr

it

ic

Ce

l

l

Subsets

1

.4

.1 P

lasmacyto

id Dendr

it

ic Ce

l

ls

Plasmacytoid dendriticcells (pDCs) are quiescentcells that are broadlydistributed in thebody. Fromtheirinitialidentificationas “lymphoblasts” by Lennert and Remmelein 1958(Lennert and Remmele, 1958), pDCs were given a number ofdifferent namesthat suggested a possible cell origin and wereexclusively based ontheirimmunophenotype (e.g., plasmacytoid T cells, plasmacytoid monocytes) (Facchetti et al., 1988). Under physiological conditions, pDCs presentplasma-likecell morphology, but onceactivated a radical morphologicaltransformation occurs;the round plasmacytoid cells acquire a

typical dendritic morphology (fromthe Greekword dendron, for’tree’) thatistightly associated with thespecialized effector functions of DC in T cell stimulation(Naik et al., 2005).

pDCs account for <0.1% of peripheral blood mononuclear cells and accumulate in inflammatorysites tocontribute to inflammatory and immune response. pDCs are

characterized as Lin− _MHC-II+_CD123(IL3R)+ _CD4+ _CD303(BDCA2)+

CD304(BD-CA4; Neuropilin-1)+ _{andare known fortheircapacityfortype 1interferonalpha}

and beta(IFN α  and β  )secretion following viralinfection bysignalingthroughthe nu- cleic acid-sensing Toll-likereceptor 7(TLR7)and TLR9(TLR7/9)(Mathan et al., 2013). They are also known fortheirantigen presentation capacity(Ginhoux et al., 2009, Sasaki et al., 2012), which is of increasing interest for anti-tumoral vaccination strategies.

They express several other characteristic markers, including sialic acid-binding immunoglobulin-likelectin H(SIGLEC-H) and bonemarrow stromal antigen 2(BST2)

in mice and blood DC antigen 2(BDCA2; also known as CLEC4C)and leukocyte

immuno-globulin-like receptor, member 4 (LILRA4; also known as ILT7)in humans

(Belz and Nutt, 2012). Other surface markers, such as maturationmarkers appear on mouse pDC as they mature in bone marrow and inthe peripheraltissues. Such markers includeLy6C (Vremec et al., 2007),Ly49Q (Toyama-Sorimachi et al., 2005), CD4, and CD8 (O’Keeffeet al., 2002). Human pDCs are usually distinguished from other DC bythelack ofexpression of CD11c(Dzioneket al., 2000).

More recently, a subset of normal pDCs wereshown toexpress CD56, a minor subpopulation of Lin− _DR+ _CD56+ _CD123+ _CD11c− _{cells were designated as pDC-}

likecells (pDLCs). Thesecells constitute 0.03% of peripheral blood mononuclearcells (PBMCs), and express BDCA2, BDCA4, and myeloid antigens, which are frequently expressed bythe malignantcounterpart of pDC (BPDCN). pDLCs exhibited modest expression ofToll-like receptors and producedlessinterferon-α  after CpG stimulation (Osakiet al., 2013).

1

.4

.2 Convent

iona

l Dendr

it

ic Ce

l

ls

Conventional DCs (cDCs) specialized for antigen processing andpresentation. Based ontheirlocalization intissuesandtheir migratory pathways, cDCs can be grouped intotwomain classes. The firstcategory ofconventional DCsis generally referredto as the migratory DCs, which develop from early precursors in the peripheral tissues, where theyact as antigen-sampling sentinels, andthenthey migratefromthe peripheral tissuestothelymph nodes. Migratory DCscan be broadlydivided into CD11b+ _DCs

(alsoknown as dermal orinterstitial DCs) andCD11b− _{DCs, which have morerecently}

been shown toexpress CD103(alsoknown as integrinαE).

Lymphoid tissue-resident DCs arethesecond majorcategory ofconventional DCs thatarefound inthe majorlymphoid organs, such as thelymph nodes,spleenand thymus. They can be classified bytheir expression ofthe surface markers; CD8α+

DCs are noted fortheir major role in primingcytotoxic CD8+_{T cell responses, CD4}+

DCs and CD4−CD8α− DCs can also present major histocompatibilitycomplex (MHC) class I-restricted antigens and moreefficiently presentMHC class II-associated antigens to CD4+ _{T cells (Belz and Nutt, 2012).}

1

.4

.3 Inf

lammatory Monocytes

During inflammation, circulating blood monocytes can be rapidly mobilized andcan differentiate intocells that possessmany prototypical features of DCs,this subset has been described as inflammatory DCs. Thesecells arisefromMDP and arerecruitedto thesite ofinflammation wheretheycan present antigensto bothCD4+and CD8+T cells (Leonet al., 2007).

Similar to macrophagesthesecells express a number ofmarkers including macrop-hage colony-stimulating factor receptor(M-CSFR)(or CD115), which is expressed

inmonocytes-macrophages andtheirprogenitors, anddrives growth anddevelopment

ofthis blood cell lineage (Bourette and Rohrschneider, 2000), and respondtogrowth factors such as granulocyte-macrophage colony-stimulating factor (GM-CSF)(Segura

et al., 2013). GM-CSF isan important hematopoietic growth factor andimmune mod-ulator, whichis producedlocally by Tcells, macrophages, endothelialcells andf ibrob-lasts upon receiving immunestimuliand it can act ina paracrine fashion torecruit circulating neutrophils, monocytes andlymphocytesto enhancetheir functionsin host defense (Shiet al., 2006).  

1

.4

.4 Langerhans Ce

l

ls

These cells areskin-residentsubset of DCsandarisefroma bone marrow-derived myelomonocytic precursorthat migratestothelymph nodesto present antigens. They sharesimilarities with macrophages intheir gene expression profilesandare deve l-opmentally dependent on signalingthroughthe macrophage colony-stimulating factor receptor(Miller et al., 2012). When activated, however, LCsacquirefeatures more inlinewith cDCs suchas their migratory ability andtheyalso sharethecDC specific transcriptionfactors (Meredithet al., 2012).

These different populations ofDC arisefrom commonlymphoid progenitor(CLP) inthe bone marrow andarerestricted to macrophage-DC progenitors (MDPs). The MDP is a common precursorthatgives riseinvivo tomonocytes, macrophages, andthe twomain subsets of DCs:cDC and pDCs(Auffray et al., 2009).

Dendritic cell development istightly controlled both bycytokines suchas FLT3 (Onaiet al., 2007), M-CSF andGM-CSF(Fancke et al., 2008) andtranscriptionfactors, which will befurther describedinthe section 1.6.

1

.5 NK

l

ineage

Natural killer (NK) cells survey host tissues for signs of infection, transformation or stress and,true totheir name, killtargetcells thathave become useless or are detr imen-taltothe host.NK cells, likecytotoxic CD8+_{T cells (alsoknown as CTLs) are defined}

based ontheir cytolytic machinery andthe killing oftheirtargetsis mediated predom-inantly via perforin and granzymes. Thesescells originate from a commonlymphoid

progenitor (CLP) and duringinfection they becomeactivated through antigen-specific receptors and by pro-inflammatorycytokines (suchas interleukin-12(IL-12) andtypeI interferons(IFNs)), and producelarge amounts ofIFNγ  (Sun and Lanier, 2011). During viralinfection, homeostasis is perturbed and matureNK cells becomeactivated, prol if-erate robustly (Raulet, 2004).cDC and pDCscontribute toNK cell-mediated protection by secreting large amounts oftypeIIFNsfollowing thetriggering ofToll-likereceptors (TLRs) and intracellular sensors of viral nucleic acids; indeed, specific pDC ablation leadsto decreasedNK cell activity (Sun and Lanier, 2011, Swieckiet al., 2010).

1

.6 Transcr

ipt

iona

l contro

l of P

lasmacyto

id Dendr

it

ic

Ce

l

l deve

lopment

Thetranscription factors (TFs) determiningthedevelopment ofDC subsetshave been

the subject of intense investigation inrecent years. The emerging paradigm isthat distinct TFs act at discrete stagesto determine or commit distinctlineage choices and cell identities(Figure 3). Sometranscription regulators such as Pu.1(Carotta et al., 2010), Gfi-1 (growth factor independent 1) a transcriptionalrepressor, and interferon regulatory factor 8 (IRF 8) are broadly required for DC development and they make their specificeffects at thelevel oftheDC progenitors. Othertranscriptionfactors such as E2-2 for example have more specificeffects withinthe pDC population (Seillet and Belz, 2013).      

1

.6

.1 Ikaros

It has been shown in numerous studiesthat Ikarosis requiredat multiple stepsinDC development andinfluences boththe earliestDC progenitors and mature DCstomain- tain pDCidentity (Allmanet al., 2006, Onaiet al., 2007). Mice expressing lowlevels of Ikaros lack peripheral pDCs,but no otherDC subsets. Loss of pDCsis associated with an inability to produce type I IFN after challenge with Toll-like receptor-7 and -9ligands.Additionally BMcells ofthesemice contain a pDC populationthat appears

FIGURE 3: Growth factors and transcriptionfactors regulating DC differentiation; the

development of both DCs and monocytes depends on high concentrations of PU.1, which regulates theexpression of thecytokine receptors FMS-related tyrosinekinase 3 (FLT3), macrophage colony-stimulating factor receptor (M-CSFR) and granulocyte- macrophage colony-stimulating factor receptor (GM-CSFR). The plasmacytoid DC (pDC) lineage requires IRF8, alowlevelof PU.1 and the absence of ID2. The differ- entiation of pDCs from an immatureprecursor requires E2-2and Ikaros, with induced loss of E2-2converting pDCs into cells that closely resemble CD8α+ _conventional

DCs. CDP, common DC progenitor; CLP, commonlymphoid progenitor; CMP, com- mon myeloid progenitor; FLT3L, FLT3 ligand; GFI1, growth factor independent 1; LMPP, lymphoid-primed multipotent progenitor; MDP, macrophage and DC progeni- tor,(From Belz and Nutt, Naturereviews, Immunology,2012)

blocked at theLy-49Q- stage ofdifferentiation andfails toterminallydifferentiate in re-sponse toFlt-3L (Allman et al., 2006). Ikaros (~50 kDa) encoded by IKZF1 (IKAROS family zincfinger 1)the gene ofsame namelocalizedto 7p12,is a member ofthe Krup-pel family of zinc fingerDNA-binding proteins andis atranscriptionfactor implicated inself-renewal activity in HSC,as well as priming of the B lymphoid transcriptional program (Nichogiannopoulou et al., 1999).Interestingly a number of these genes are also expressed in pDC. The precisedownstream targetsforIkaroshave not beenidenti- fied inthe pDClineage.

1

.6

.2

E2-2

As discussed above, pDCsarea distinctlineageseparatedfromthecDCs, at a late development stage, bydifferential regulation of discrete geneexpression programs by distincttranscription factors, inparticular E2-2 (Cisse et al., 2008, Ghoshet al., 2010). E2-2 (~71k Da), encoded by Tcf4,located on 18q21locus,is a member of E-box (or E) proteinsthat belongtothe basic helix-loop-helix(bHLH)transcriptionfactor family.

The members ofthis family bindto regulatory E-box sequences (CATATG/CACCTG)

on target genesand activate transcription byformingeither heterodimers or homod-imers(Kee, 2009). E2-2 is highlyexpressed in pDC,at lowerlevelsin Bcells, andis almost undetectable in otherimmunecells. pDCs arevery sensitive toE2-2expression; inE2-2+/− _{mice as in human patientswith Pitt-Hopkins syndrome whohave E2-2hap-}

loinsufficiency, pDC numbers and IFN-α  secretion are profoundly impaired (Cisse et al., 2008).

E2-2 binds directly tothe promoter regions of Spi-B, Irf8, and Irf7 genes, whichare important factors for pDC development (Cisse et al., 2008, Ghosh et al., 2010).Induced deletion of E2-2 in mature pDC is associated with the loss of pDCs markers, increased MHC class IIexpression and Tcell primingcapacity, acquisition of dendritic morphology, and cDC phenotypeaccompanied byincreased expression of Id2 (an antagonist E protein, which is abundantly expressed in all cDC) and the loss of pDClineageidentity (Ghoshet al., 2010).

1

.6

.3 SPI-B

SPI-B (~29 kDa), is encoded fromSPI-B genelocated at 19q13locus, andisan ETS-

domain transcription-factor family member and is closely related to PU.1, with 70% sequenceidentity intheir DNA-binding domain(DBD)(Chrast et al., 1995).SPI-B- deficient pDCs present defects in TLR7/9-inducedtypeIIFN production, Sasaki et al. demonstrated that pDCs were decreasedin BM ofSpi-B−/− mice whereas per iph-eral pDCs wereincreased;this uneven distribution suggestedthedefective retention of

function and development in mice (Sasakiet al., 2012). Knockdown experiments in CD34+ _{precursor cells demonstratedthatthe ability ofthesecells todevelop pDCsin}

both invitro and invivo assays was stronglyinhibited(Schotteet al., 2004).

1

.6

.4 Interferon regu

latory factor 8

Interferon regulatoryfactor 8 (IRF8) (~48 kDa) belongs to the IRF transcription fac- tor family. IRF8 encoding gene is locatedat 16q24.1 locus andis particularly highly

expressed in common dendriticcell precursor (CDP), pDCs, and CD8a DCs(Becker

et al., 2012).

Thefunction of IRF8inDC development appearsto beconserved across humans

and mice. In vivo experiments have shown thatIRF8 plays an important roleinregu- lation of proliferation anddifferentiation of hematopoietic progenitorcells. IRF8 was recentlyfound to promote CDP development andtheinitialcommitment tothe DC lineage by inhibiting granulopoiesis (Becker et al., 2012). pDCs and CD8a DCs help to establishantiviral and antimicrobial responses by producing type I IFN and IL-12, respectively. These functions are impairedinIrf8−/− _{mice, which are more suscept}

i-ble to viral infection, a phenotype shared by the human patientswith IRF8 mutations (Hambletonet al., 2011).      

1

.6

.5 RUNX2

The Runx family oftranscriptionfactorsinvertebrates consists ofthree proteins (Runx1-3). Runx1 (~55 kDa) and Runx3 (~48 kDa) have been extensively studied within the hematopoietic system, where they often act astranscriptionalregulators that direct cell fate choices. Runx1 and Runx3 are critical regulators of T cell differentiation but Runx2 has beenspecificallyidentified as a master regulator of bone development. It has beenrecently shown that pDCsspecifically express Runx2(~57 kDa) in an E2- 2–dependent manner. Runx2,locatedat 6p21locus, isrequiredforthe expression of several pDC-enriched genes, including the chemokine receptors Ccr2 and Ccr5. Ma- ture pDCsexpress highlevels of Ccr5 atthe cellsurface, and Ccr5-deficient pDCs are

greatly reduced inthe peripheryrelativetothe bonemarrow. pDCsinRunx2-deficient mice developed normallyinthe bonemarrow but are greatly reducedinthe periphery (caused by the retention of mature Ly49Q+_{pDCs in the bone marrow). These data}

indicate that Runx2 facilitates pDC homing to peripheral lymphoid organs in a Ccr5 dependent manner (Sawaiet al., 2013).

1

.6

.6 L-Myc

MYC genes encode nuclear factors that criticallyregulate stem cell and metabolic func-tions,cell proliferation and apoptosis. Theirderegulation contributes tothe genesis of many humantumors. Inmammals, there are fourrelated genesinthefamily, encoding

4 factors: c-MYC (~67 kDa), N-MYC(~49 kDa), L-MYC(~40 kDa)and S-MYC

(~50 kDa).It is very well known now that the enhanced expression of Myc proteins contributes to almostevery aspect oftumorcell biology.

The c-MYC gene islocatedat 8q24.21locus andisthe most studied member of thisfamily.Deregulated expression ofc-MYC occurs in a broadrange of human cancers and is often associatedwith poor prognosis,indicating akey roleforthis oncogeneintu- mor progression. c-MYC regulates cellular proliferation, metabolismand maintenance of progenitor populations. MYC proteins bind cognateDNA sequencesin cooperation with DNA binding partners (MAX)through aconserved basic helixloop helixleucine zipper. At least c-MYC has been shown to directly regulate thetranscription of numerous genes but also tocontrolthespeed at whichtranscription proceeds. The third member of myc family of proto-oncogenes is L-MYC, which islocated at

1p34.2 locus and is also required for normal embryonic development. Forced

expression of L-Mycexerts weaker effects than c-MYC forcell growth, apoptosis and transformation (Wasylishen et al., 2011) but is more efficient inreprogramming fibroblaststowardsinduced pluripotentstem cells (Nakagawa et al., 2010).

Interestingly it has been recently shown that L-Myc is selectively expressed in pDCs and itsexpression is initiatedin commonDC progenitor. Theknockout experi- ments inmice have shown thattheloss ofL-Myc in DCs causes a significant decrease

ininvivo T-cell priming duringinfection. It has beenshown likewisethat some subsets of DCs such as migratory CD103+ _{conventional DCsinthelung andliver are greatly}

reduced at steady state, whenc-Myc is replaced byL-Myc inimmatureDCs (Kc et al.,

Chapter 2

B

last

ic

p

lasmacyto

id

dendr

it

ic

ce

l

l

neop

lasm

2

.1 C

lass

if

icat

ion

and

c

l

in

ica

l

features

Others and weidentifiedleukemic derivatives of pDC inthe early 2000s(Chaperot et al., 2001).Following on fromthis,intense clinical, pathology analysis and bio- logical and functional investigations led to pDC-derived leukemic disorders being classified as a new entity blastic plasmacytoid dendritic cell neoplasm(BPDCN)(formerly called CD4+_/CD56+ _{hematodermic neoplasm or plasmacytoid dendritic cell leukemia)}

within AML, inthe currentWHO classification of hematological malignancies. Achar- acteristic of BPDCN is a clinicallyaggressive coursewith a mediansurvival of 12-14 months irrespective oftheinitial pattern of disease(Fachetti, 2008). BPDCNis a rare disease representing <1% of acute leukemiacases (Jacob et al., 2003), predominantly affects males, with asex ratio of 3:1,with an average of 67 years old, although some pediatric cases have also been reported (Feuillardet al., 2002). Thisaggressive disease presents a milder clinical course in childrenthanin adults(Jegalian et al., 2010).

Clinical presentation, consisting at thetime of diagnosis of a solitary cutaneous lesionthatrapidly disseminatesin multiplessites includinglymph nodes, bonemarrow, blood or central nervous system(CNS)(Jardinet al., 2011). BPDCN cutaneous nodules

are present in 90% of patients (the most common feature; including nodules, plaques, or bruise-like areas)with or without bonemarrow, peripheral blood(PB)(60-90%) and

lymph nodes (40-50%)involvement. Lymphadenopathy orspleen enlargement or both

and cytopenia isfrequent(Fachetti, 2008).

Morphologically BPDCN is usually pleomorphic with cell size varying from small tolarge,with irregular nuclei, fine chromatin and onetoseveral small nucleoli. The cytoplasm is non-granularbut displays a heterogeneous structure and cytoplasmic membrane often exhibits pseudopods(Fachetti, 2008, Garnache-Ottouet al., 2007).

As mentioned before,leukemicplasmacytoid dendriticcells have beenclassified as blastic NK-cell, lymphoma/leukemia oragranular CD4+_{, CD56}+ _{and/or hemato-}

dermic neoplasm of myeloid precursors(Brody et al., 1995, Facchetti et al., 1990). Chaperot and colleagues werethefirst toreport andrecognizethatthis diseaseis ama-

lignancy ofthe pDClineage (Chaperotet al., 2001). They performedan

immunophe-notypic study on the leukemic cells of 7 patients (6 men and 1 woman from 8 to 86 years of age, mean = 63 years) andidentifiedthatthese cells presentedexpression of

the following markers: CD4+_{, CD56}+_{, CD3}−_{, CD13}−_{, CD33}−_{,and CD19}−_{. They}

completedthesefindings with functional studies and demonstratedthat,as reportedfor normal pDCs, inthesame culture conditions,leukemiccells becamepowerful inducers of T-cell proliferation.In additiontumor cells from peripheral blood andlymph nodes of two of these patients produced IFN-α  in response toinactivated influenza virus, a distinctfeature of pDCs and as suchreferredtoas “naturalinterferon-producingcells”. On the basis ofthese observations Chaperot et al. proposedthat BPDCN is derived froma plasmacytoid dendritic cell interleukin-3α  (IL-3Rα/CD123+) precursor.

Be-sides they confirmedthatthesecells sharedalso somemarkers with T (CD4) andNK

cells (CD56), suggesting a relationtolymphoidlineage origin ofthese neoplasticcells (Chaperotet al., 2001,Fachetti, 2008).

Jacob et al. have documented, throughan analysis of publishedcases in li tera-turethatan important number of malignantcases showing similar featuresas explained

for BPDCN. They assessed thentheexpression of more specific pDC-relatedmarkers

BDCA2+ or BDCA4+ (Jacob et al., 2003). Inaddition, inastudy of 45 cases of BPDCN neoplasmsandtheir potentialcutaneous mimics, someimmunohistochem i-cal markers, such as SPIB, BDCA-4, IRF-8, BCL11A and CD2AP, have been reported as toolsfor BPDCN diagnosis(Montes-Morenoet al., 2013).

To datethe etiology of BPDCNis unclearbut its associationwith myelodysplasia in somecases may suggest arelatedpathogenesis (Fachetti, 2008). It has beenreported thatsome BPDCN patients presenteda prior history of myelodysplastic syndromes (MDS) (Feuillard et al.,2002). MDS is a clonal disorder of hematopoieticstem cells characterized by ineffective hematopoiesis, morphologic dysplasia, peripheral blood cytopenias, and propensitytotransformationto acute myeloidleukemia (AML) (Mufti et al., 2003). The geneticfeatures ofMDS will be discussedinthefollowing section.

BPDCN, categorized previously as atype ofNK-cell lymphoma,isnow classified as arare subgroup ofAML andrelated precursor myeloid neoplasmsinthe 2008World Health Organization (WHO) classification 4th edition nomenclature, (Fachetti, 2008) becausethe origin ofthe malignancy has beenidentifiedas precursor of pDCs(Chaperot et al., 2001).

2

.2 Mo

lecu

lar

pathogenes

is

of

BPDCN

2

.2

.1 Chromosoma

l aberrat

ions

In 2002 Leroux et al. have documentedforthefirst timethe cytogenetic features of BPDCN, by using conventional and Fluorescence in situ hybridization (FISH) on 18 adults and 3 children with CD4+_{, CD56}+ _{DC2 acute leukemia. Clonal, mostlycomplex}

chromosome aberrations werefound in 14 patients(66%). The aberrantcytogenetic fea- tures representedimportant genomicimbalances, a combination/accumulation of chro-mosomal anomalies and bothlymphoid and myeloid lineage-associated rearrangements (Leroux et al., 2002, Sapienzaet al., 2014). Thisissuggestive ofcomplex multistep tu- morigenic mechanisms and issupportive ofthehypothesis that BPDCNmay arisefrom an undifferentiated progenitor (Leroux et al., 2002). Leroux et al. showed that two

thirds of patientswith BPDCN presentedcomplex karyotypes and documentedsix ma- jor recurrent specific chromosomal abnormalities; 5q21 or 5q34 (72%), 12p13 (64%),

13q13-21 (64%), 6q23-qter (50%), 15q (43%) andloss of chromosome 9 (28%). Re-

current deletions of chromosomal regions suchas 4(4q34), 9(9p13-p11 and 9q12-q34) and 13(13q12-q31) have beenalso observed (Lerouxet al., 2002).

Inaddition to FISH, otherapproaches have been employed tocharacterizethe geneticlandscape of BPDCN. For example, comparative genomichybridization (CGH) array studies in skin biopsies from 11 cases (5 cases with BPDCN and 6 cases with AML) coupledto Geneexpression profiling(GEP), has demonstratedrecurrent deletion

of regions onchromosome 4(4q34),chromosome 9(9p13-p11and 9q12-q34),and

chromosome 13 (13q12-q31) are recurrent in BPDCN (Dijkmanet al., 2007) that are

distinctive of BPDCN comparedtoAML with cutaneouslesions. BPDCNwas found

toshow a distinct geneexpression profile comprisingexpression of both myeloid and

lymphoid genes comparedtoAML.

In another study Jardinet al. used array CGHto delineatenovel candidateregions and disease-related genes in BPDCN.They studied 9cases inwhichthe most frequent findings were the loss of chromosome 9 (four cases) orchromosome 13(five cases). The most frequent partial chromosomallosses affected 5q (four cases), 7p(two cases), 8q (two cases), 9p(two cases), 12p (six cases), 13q(two cases) and 17p (threecases) (Jardinet al., 2009).

Taken togethertheseobservations demonstratedthat 5q deletionis afrequent phe-nomenon in BPDCN anditis mostly associatedwith “lymphoid type” anomalies (up to 70% ofcases with abnormal karyotypes)inthis disease (Jardinet al., 2009, Leroux et al., 2002, Petrellaet al., 2005).

It iswell known that deletionsinvolving the long arm of chromosome 5 (chr5), del(5q), arethe most common cytogenetic abnormalitiesinMDS andsecondary AML. Thesechromosomallossesareespecially prevalent in AML arising after prodromal MDS or in MDS and AML arising after previous cancertreatment with alkylating agents or radiotherapy (approximately 40% of patients)(Ebert, 2009).

As mentioned briefly before,MDS comprises a heterogeneousspectrum of myeloid malignancies, andthefrequent cytogenetic alterations inthis disease consists of (-5/5q-,-7/7q-, +8, 20q-, and-Y) (Haase, 2008). Besides 5q-syndrome, which isa hematologic disorder with isolated 5q deletion and blast count of <5% that is assoc i-ated with specific MDS clinical features andfavorable prognosis (Ebert, 2009) alarge proportion of del(5q) patients have bone marrow (BM) blasts exceeding 5%, complex cytogenetic, and extremely poor prognosis andarerefractory to available treatments (Byrdet al., 2002, Giagounidiset al., 2006,Grimwade et al., 2001).

Chromosome 5qalterations have already beenassociated to haploinsufficiency mechanisms caused bycritical 5q target gene(s)that incooperation with additional signaling net- works drives malignanttransformationand clonal evolution inthese disorders(Ebert, 2011).It has recently been demonstrated that the deletion of m iR-146a in del(5q) MD- S/AML cells is associated with increasing cell survival and proliferation of the propa- gating cells troughtheTRAF6/p62/NF-κB complex (Fang et al., 2014).

Untilthe present work, no study has addressedidentification of 5qdriver genes in BPDCN. A significant proportion of BPDCNcases with normalkaryotypes are also described but have notundergone detailed molecularinvestigations either by molecular cytogenetic, sequencing orGEP.

2

.2

.2 Recurrent gene mutat

ions in

BPDCN

2.2.2.1 Tumor suppressor genes

Despitethe findings on ontogenetic origin of BPDCNtumors, the genetic causes and oncogenic signaling events involved inthe malignant transformation ofthis disease are still unknown. Preliminary results from molecular studies suggestedthe presence of abnormalities in genes known toconfera poor prognosis in other hematopoietic malignancies. BPDCN cells display recurrent alterationsleadingto a combination of deletions of several tumor suppressor genesincluding: RB1, CDKN1B, CDKN2A, or

TP53 (Jardin et al., 2009),indicating that deletion events altering G1/Sregulation are

crucialfor BPDCN oncogenesis.

      TP53  

TP53 (tumorprotein 53orthe human p53 gene) is one ofthe most commonly mutated

tumor suppressor genesin humancancer. Many ofthesemutations leadtoexpression

of mutant p53 proteins that dominantly interactwith normal p53activity and thereby drive genomeinstability, abnormal proliferation andcell survival and treatment res is-tance (Muller and Vousden, 2013). It has been shown that tp53 inactivation strongly cooperates with oncogenicKras to induceaggressive AML: myeloid progenitorcells expressing oncogenic Kras and lacking p53 become leukemia-initiating cells, resem-bling cancer stem cells capable of maintainingAML invivo. These results suggested the ability oftp53tolimit aberrantself-renewal andcontribute toitstumor suppressor activity (Zhaoet al., 2010). In orderto determinethemutational statusin BPDCN Jardin et al. performed polymerase chainreaction (PCR) assaysfollowed by direct sequencing on thirteen patients with BPDCN. Five cases displayed TP53 mutations (38%), leading to changesinits functional domain (Jardinet al., 2011), indicating thatTP53 mutation isarecurrent genetic anomaly in BPDCN.

RB protein families  

The retinoblastoma (RB) gene family iscomposed of RB1, located at chromosome 13q14.2, and two other RB-related genes designated as retinoblastoma-like 1 (RBL1) and retinoblastoma-like 2 (RBL2) located respectively at chromosome 20q11.2, a re-gion of specialinterest because of its associationwith some myeloid disorders(Claudio et al., 2002) andchromosome 16q12.2in which deletions orloss of heterozygosityhave been reported inseveral humanneoplasms respectively (Di Fioreet al., 2013). These evolutionary conserved genes are structurally related, and play a key roleinregulating advancement ofthe cell division cycle fromthe G1 to S-phases. They regulate neg- atively cell cycle entry, by controlling E2Ftranscriptionfactors andcyclin dependent

kinases. Stimulation of cell cycle entry bygrowth factor signaling leadstoactivation of cyclin dependent kinases,resulting in phosphorylation and inactivation of the RB family proteins, and finally E2Factivation andDNA synthesis (Figure 4)(Henley and Dick, 2012).

FIGURE4: The central role of the retinoblastoma protein (pRb) in cell-cycle pro-

gression. Entranceinto S phaseis characterized by phosphorylation of pRb by cyclin D1–cyclin-dependent kinase (CDK)-4, cyclin D1–CDK6 and cyclin E–CDK2 com- plexes. The phosphorylation of pRb is associated with release of the E2F1–3 tran- scription factors, which then activategenes thatare required for cell-cycle progression. From (Hilary A. Coller, Nat RevMol Cell Biol, 2007)

It has been demonstratedthat mutational inactivation of RB1 causesthe pedi- atric cancer retinoblastoma, while deregulation ofthepathway in whichit functionsis common in most types of human cancer(Goodrich, 2006). Inactivation ofRB tumor- suppressor gene has been already reported inlymphoid malignancies(Hangaishi et al., 1996). By using Gene expression profile (GEP) on 11cases (5cases with BPDCN and 6 cases with AML), Djikman et al. demonstrated thatthereis a diminishedexpression

inoncosuppressor genes suchas RB1 and LATS2.LATS2 isalso an importantfactor for

implicatedincell cycle control(Dijkmanet al., 2007).

In orderto establishcorrelationswith disease outcome and bycombining clin ico-pathologic findings and genetic data, Lucioni et al. utilized array CGH on 21cases of

detected a common deleted region (CDR) on 13q13.1-q14.3 involving RB1 (Lucioni et al., 2011).

2.2.2.2 Gene mutations affectingcell signaling FLT3

The FMS-liketyrosine kinase 3(FLT3) gene,located on 13q12, encodes a membrane- boundreceptortyrosine kinase(~140 kDa)that has a crucialrole in normal hematopo i-esis. It containsfive extracellular immunoglobulin-like domains(E), atransmembrane

domain(TM), ajuxtamembrane domain(JM) and twotyrosine-kinasedomains (K)that

are linkedthroughthetyrosine-kinaseinsert(KI) (Figure 5). CytoplasmicFLT3 under- goes glycosylation, which promotes itslocalizationtothe membrane, whereit remains as a monomeric, inactive protein onthecell surface until FLT3 ligand (L) bindsthe receptor andinducesreceptor dimerization. Itsdimerization promotes phosphorylation (P) ofthetyrosine-kinase domains,thereby activating the receptor anddownstream ef- fectors. FLT3 is normally onlyexpressed inprimitive hematopoietic precursors within the bonemarrow (Stirewalt and Radich, 2003).

Clinical and experimental evidence bothindicate thatFLT3 is a proto-oncogene with the capacityto enhancesurvival andproliferation ofleukemia blastcells (Stirewalt and Radich, 2003). There are two broadcategories ofactivating mutations ofFLT3. The JM domain ofFLT3 oftenhas an autoinhibitory function, therefore mutations disrup t-ingthis domain and causingthe deformation ofthe secondarystructure ofFLT3, result inconstitutive activation of the tyrosine kinase domain of this receptor (theso-called FLT3-ITD “internaltandem duplication mutations”). Other major FLT3-activating mutationconsists of point mutations withinthetyrosine kinase domain(TKD), pa r-ticularly onthe activation loop at aspartate 835(D835) (Grunwald and Levis, 2013) (Figure 5). The constitutive activation ofFLT3 promotesligand-independentcell pro-liferation and blocks myeloiddifferentiation of early hematopoieticcells (Stirewalt and Radich, 2003). Mutations intheFMS-liketyrosinekinase gene characterize morethan 30 % of AML cases. FLT3 internaltandem duplication(ITD)mutations, accountingfor

FIGURE 5: A schematic diagram of the FLT3 receptor tyrosine kinase showing the

locationof theinternaltandemduplication of genes within thejuxtamembranedomain and point mutations and gene insertionsinthesecond kinase domain. Illustration by Kenneth Probst. From (Mark R. Litzow, Blood, 2005)

approximately 23 % of AML cases, are associatedwith a particularly poor prognosis

(Grunwald andLevis, 2013).

Pagano et al. have done molecularanalysis on 14 BPDCN patientsandfound thatthree(21%) ofthese patients hadFLT3-ITD mutations (Pagano et al., 2013). Gene-expression profiling on 8 BPDCN patient biopsiesalso demonstratedthatthereis a high expression ofFLT3 in some ofthese samples, anditwas paradoxically associatedwith monoallelic deletion ofthelocus containingthis proto-oncogene(Dijkman et al., 2007).

      BRAF  

The RAF-MEK-ERK signaltransductionpathway is aconserved RAS-activated protein

kinase cascadethat regulates cell growth, proliferation, anddifferentiation in response to growth factors, cytokines, and hormones with RAF as thefirst effector identified

RAF activation initiates by RAS-GTP associationwith the N-terminal RAS b ind-ing domain(RBD), whichis associated by concomitant conformational changes and re-cruitment tothecell membrane, promoting changesin RAF phosphorylation, stimula t-ingits serine/threonine kinaseactivity. Thisresultsinthetriggering sequential phospho-rylation and activation ofMEK andextracellular signal-regulated kinase (ERK)(Wan et al., 2004).Constitutive activation of the RAS-ERK signalingpathway is common to numerous cancers. Almost 90% of BRAF mutations result in a Val600_{Glu (V600E)}

amino acid substitutionleadingtoconstitutive kinaseactivation (Hoeflichet al., 2006).

Recently Go et al. examined BRAF mutation by using Sangersequencingand peptide nucleic acid clamp real-time polymerase chainreaction(PNAcqPCR) in a series of 129 cases of histiocytic, dendritic cell and otherrelatedlesions. PNAcqPCR isa sensitive method for detection of mutationsin the presence of alargeexcess ofwild- typeDNA. PNA is asynthetic DNA analogue,in whichtheribose/phosphate backbone ofthe DNA has beenreplaced by N-(2-aminoethyl)-glycine units linked by peptide bonds. PNA bindsstrongly to complementary DNA by Watson-Crick base pairing, whereas one single mismatch will severely destabilizethecomplex.

This analysis detectedBRAF V600E mutations, and Histiocytic sarcomaexhib- itedthe highest rate ofBRAFV 600E_{(62.5%, five of eight),followed by Langerhanscell}

tumors (25%, seven of 28),follicular dendritic cell sarcoma (18.5%, five of 27) and giant cell tumor(6.7%,two of 30)(Goet al., 2014).

2.2.2.3 Transcriptionfactors

Ikaros

By using a whole-exome sequencing(WES), Menesezet al. provide thefirst mutational profiling of BPDCN. They identified recurring mutations for a group ofnewly ident i-fied genes, suchas IKZF3 (Ikaros family of zinc-finger protein hematopoietic-specific transcriptionfactors involvedintheregulation oflymphocytedevelopment) andZEB2

(Zincfinger E-box-binding homeobox 2,is atranscriptionfactor implicatedindevelop- ment), with prevalence range of 12-16% in BPDCN. Itis of notethatIkaros mutations are also describedin B-ALL (Kastneret al., 2013, Menezeset al., 2014).

2.2.2.4 Epigenetic regulators

Beside recurrent mutations that confer a growth advantage by activating downstream effectors of various signaling pathways and mutationsthatimpactthe expression of key transcriptionaltargetsin myelopoiesis,emerging genetic data suggestthatthere are additional classes of mutationsthat commonly occurin patientswith myeloid mal ignan-cies. The most prominentexamples are from recent genome-wide and candidate-gene studies that have identified somatic alterations in genes that encode proteins regulat- ingDNA methylation and post-translational histone modifications. These data suggest that somatic alterationsin epigeneticregulators are a common geneticevent in cancer including myeloid malignancies andcontribute to hematopoietictransformation (Shih et al., 2012).

Inthe following section I brieflyintroduce epigeneticsderegulation, whichis an importantemerging hallmark ofcancer, and will discuss specificallythe mutationsaf- fecting epigenetic factors thathave been demonstratedto be associatedto BPDCN and similar hematological malignancies.

Ep

igenet

ics:

an

emerg

ing

ha

l

lmark

of

cancer

As normalcells evolve progressively to a neoplasticstate, they acquire a succession of capabilitiesthat helpthem to becometumorigenic and ultimately malignant.Tumors are complex tissues composed of multiple distinctcell typesthatinteractwith oneanother, andtumor microenvironment contributes actively totheirdevelopment. As it proposed by Hanahan and Weinberg (Hanahan andWeinberg, 2011),the hallmarks of cancer con-sist ofthefollowing biological capabilities; sustainingproliferative signaling,evading growth suppressors,resistingcell death, enablingreplicative immortality, genomeins ta-bility and mutation, tumor-promoting inflammation,inducing angiogenesis,activating

invasionand metastasis. They have also proposed two emerging hallmarks based on recent research, comprising; the deregulation of cellularenergetics and the capability of avoiding immune destructionintheirlast model(Figure 6).

FIGURE 6: The hallmarks of cancer.From (Hanahan and Weinberg, Cell, 2011)

Inthe past decade aremarkable acceleration has been madeinthevalidation ofthe concept as well as genetic anomalies that result in tumor formation and progression, canceris an epigenetic disease (Baylin and Jones, 2011). Following thediscoveries of DNA methylation abnormalities,learning about chromatin covalent modifications,their organization andrelevanceto geneexpression, brought ustotheemerging view of what is now called “the cancer epigenome”. This comprises alterations in epigenetic information as opposed to DNA sequence changes. By definition epigenetic alterations are heritable and somatically reversible making themideal targetsfor treatmentinterventionincancer.

General overview on epigenetics

The geneticcoderesides withina negatively charged DNA polymer. The binding of theDNA fiber around basic histone proteinsresultsin neutralizingthenegative charges

and allows theDNA to befoldedinto chromosomes and compacted upto 10,000times withinthe nucleus. Wrapping DNA around histonecore, which iscomposed of two copies of each ofthe histone proteins H2A, H2B,H3 and H4,formsthe most basic unit of eukaryotic chromatin:the nucleosome (Dawson andKouzarides, 2012)(Figure 7).

Histones are not merely DNA-packaging proteins, but molecular structures that participate inthe regulation of gene expression. They storeepigeneticinformation through post-translational modifications ontheir amino-terminal histone“tails”, as ly- sineacetylation, arginine andlysine methylation, andserine phosphorylation. These modifications affect genetranscription andDNA repair(Jiang and Pugh, 2009).

FIGURE 7: Overall structure of the epigenome in human cells. From (Baylin and