Plenary Paper

Clonal architecture of chronic myelomonocytic leukemiasRaphael Itzykson,1,2,3 Olivier Kosmider,4,5,6 Aline Renneville,7,8,9 Margot Morabito,1,2,3 Claude Preudhomme,7,8,9

Celine Berthon,8,10 Lionel Ades,11 Pierre Fenaux,11 Uwe Platzbecker,12 Olivier Gagey,13 Philippe Rameau,2

Guillaume Meurice,2 Cedric Orear,2 Francois Delhommeau,1,2,14,15 Olivier A. Bernard,2,16 Michaela Fontenay,4,5,6

William Vainchenker,1,2,3 Nathalie Droin,1,2,3 and Eric Solary1,2,3

1Inserm UMR1009, and 2IFR54, Institut Gustave Roussy, Villejuif, France; 3Faculty of Medicine, University Paris-Sud, Le Kremlin-Bicetre, France; 4Assistance

Publique–Hopitaux de Paris, Hematologie Biologique, Hopital Cochin, Paris, France; 5Faculte de Medecine, Universite Paris Descartes, Paris, France;6Departement d’Immuno-Hematologie, Institut Cochin, Paris, France; 7Laboratory of Hematology, Biology and Pathology Center, CHRU of Lille, Lille, France;8University of Lille Nord de France, Lille, France; 9Inserm, U837, Team 3, Cancer Research Institute of Lille, Lille, France; 10CHRU de Lille, Clinical hematology

Unit, CHU Lille; 11Hopital Avicenne, Assistance Publique-Hopitaux de Paris, Service d’Hematologie Clinique, University Paris XIII, Bobigny, France;12Medizinische Klinik und Poliklinik I, Universitatsklinikum ‘Carl Gustav Carus’ Dresden, 01307 Dresden, Germany; 13Assistance Publique–Hopitaux de Paris,

Service de Chirurgie Orthopedique, Le Kremlin-Bicetre, France; 14UPMC Univ Paris 06, GRC n˚07, Groupe de Recherche Clinique sur les Myeloproliferations

Aigues et Chroniques MyPAC, F-75012, Paris, France; 15Assistance Publique–Hopitaux de Paris, Hematologie et Immunologie Biologique, Hopital Saint-

Antoine, Paris, France; and 16Inserm UMR985, Institut Gustave Roussy, Villejuif, France

Key Points

• Early clonal dominancemay distinguish chronicmyelomonocytic leukemiafrom other chronic myeloidneoplasms with similar genemutations.

• Early dominance of TET2-mutated cells in thehematopoietic tissuepromotes myeloiddifferentiation skewingtoward the granulomonocyticline.

Genomic studies in chronic myeloid malignancies, including myeloproliferative

neoplasms (MPN), myelodysplastic syndromes (MDS), and MPN/MDS, have identi-

fied common mutations in genes encoding signaling, epigenetic, transcription, and

splicing factors. In the present study, we interrogated the clonal architecture by

mutation-specific discrimination analysis of single-cell–derived colonies in 28 patients

with chronic myelomonocytic leukemias (CMML), the most frequent MPN/MDS. This

analysis reveals a linear acquisition of the studied mutations with limited branching

through loss of heterozygosity. Serial analysis of untreated and treated samples

demonstrates a dynamic architecture on which most current therapeutic approaches

have limited effects. The main disease characteristics are early clonal dominance,

arising at the CD341/CD382 stage of hematopoiesis, and granulomonocytic differen-

tiation skewing of multipotent and common myeloid progenitors. Comparison of

clonal expansions of TET2 mutations in MDS, MPN, and CMML, together with

functional invalidation of TET2 in sorted progenitors, suggests a causative link

between early clonal dominance and skewed granulomonocytic differentiation.

Altogether, early clonal dominance may distinguish CMML from other chronic myeloid

neoplasms with similar gene mutations. (Blood. 2013;121(12):2186-2198)

Introduction

Hematologic malignancies result from evolutionary processes drivenby stepwise accumulation of genetic mutations with clonalexpansion and selection.1 This model has been refined with theobservation of complex, highly subclonal architectures2 and thefinding of simultaneous acquisition of genetic lesions.3 Explorationof clonal evolution uses various methods, including deep sequencingof bulk tumors4 and sequencing of single cells or colonies,5 andneeds to integrate the tissue-specific cellular hierarchy to understandwhy different neoplasms emerging from the same tissue and sharinggenetic lesions harbor distinct phenotypes and prognoses. Becauseof the number of associated gene mutations,6,7 chronic myelomo-nocytic leukemia (CMML) is a unique model to address clonal archi-tecture in chronic myeloid malignancies.

The characteristic feature of CMML is the accumulation ofmonocytes, together with dysplastic granulocytes, in the peripheral

blood, bone marrow (BM), and spleen.8 Mutated genes encodesignaling molecules (NRAS, KRAS, CBL, JAK2), transcriptionfactors such as RUNX1, epigenetic regulators (TET2, ASXL1,EZH2, UTX, IDH1, IDH2, DNMT3A), and splicing factors (SF3B1,SRSF2, ZRSF2, U2AF35).6,7,9-15 None of these mutations isspecific of the disease, as they were also identified in MPN16 and inMDS.17 In MPN, mutations in epigenetic regulators such as TET2or ASXL1 can be either late events associated with diseaseprogression, or early events that precede mutations in signalingmolecules.5,18,19 In MDS, cell dysplasia in diverse myeloid line-age results from various combinations of gene mutations,17 andknowledge on genotype and phenotype associations is currentlylimited.20 The role and place of individual mutations regardingclonal emergence and progression of chronic myeloid malignanciesremain poorly understood.

Submitted June 28, 2012; accepted November 28, 2012. Prepublished online

as Blood First Edition paper, January 14, 2013; DOI 10.1182/blood-2012-06-

440347.

The online version of this article contains a data supplement.

The publication costs of this article were defrayed in part by page charge

payment. Therefore, and solely to indicate this fact, this article is hereby

marked “advertisement” in accordance with 18 USC section 1734.

© 2013 by The American Society of Hematology

2186 BLOOD, 21 MARCH 2013 x VOLUME 121, NUMBER 12

For personal use only.on May 30, 2017. by guest www.bloodjournal.orgFrom

The prognosis of CMML is poor, and aside for the fewpatients eligible for allogeneic stem cell transplantation, there isno curative treatment. Therapeutic options include hydroxyureato control myeloproliferation, erythropoiesis-stimulating agentsto correct anemia, and hypomethylating agents to delay pro-gression.8,21 Novel therapies, such as mitogen-activated proteinkinase inhibitors, are in development. The total number ofmutated genes seems to be a strong prognostic factor in CMML,7

suggesting a key role for clonal complexity in leukemogenesisand resistance to therapy.

Here, we explored the clonal architecture of 28 CMML bymutation-specific discrimination analysis of single-cell–derivedclones. The genetic classification of individual cells allowed adesignation of subclones and the assembly of putative evolu-tionary trees. We established the sequence of mutation acquisi-tion, the dynamics of clonal expansion during hematopoieticdifferentiation, and its relationship to the disease phenotype andevolution, both in untreated and in treated patients. Our resultssuggest that early clonal dominance may specify CMML amongother chronic myeloid neoplasms.

Patients and methods

Patients

Blood and BM samples from patients with CMML from GroupeFrancophone des Myelodysplasies (study group is described in supple-mental Methods) and control BM samples from older (.50 years)patients undergoing hip surgery were prospectively collected afterinformed consent according to the Declaration of Helsinki. Details areprovided in supplemental Methods. The noninterventional study anddecitabine trials were approved by the Cochin Hospital and Aulnay-sous-Bois ethic committees (Comite Consultatif de Protection des Personnes),respectively.

Flow cytometry and cell sorting or cloning

After immunomagnetic enrichment, BM CD341 cells were sorted inthe following fractions: CD341CD382CD901 (hematopoietic stemcells [HSC]), CD341CD382CD902 (multipotent progenitors [MPP]),CD341CD381CD45RA2CD1231 (common myeloid progenitors[CMP]), CD341CD381CD45RA1CD1231 (granulocyte-monocyte pro-genitors [GMP]), and CD341CD381CD45RA2CD1232 (megakaryocyte-erythrocyte progenitors [MEP]). Peripheral blood CD341 cells weresorted as CD341CD382 and CD341CD381 fractions. Sorted fractionswere cloned at 1 cell per well in 96-well plates. Details are provided insupplemental Methods.

Gene mutation analysis

DNA extracted from CD141 and CD31 sorted cells (Norgen Biotek,Thorold, ON, Canada) was submitted to whole-genome amplification(Repli-G; QIAGEN, Hilden, Germany) for gene mutation screening ofFLT3 (internal tandem duplications and tyrosine kinase domain mutations),NPM1 (exon 12), JAK2 (V617F), KIT (exon 17), DNMT3A (exon 23),IDH1 (exon 4), IDH2 (exon 4), TET2 (exons 3-11), CBL (exons 8-9),RUNX1 (exons 3-8), ASXL1 (exon 12), EZH2 (exons 2-20), SF3B1 (exons13-16), U2AF35 (exons 2 and 6), ZRSR2 (exons 1-11), and SRSF2 (exon 2).All abnormalities were validated on nonamplified DNA. Details areprovided in supplemental Methods.

Whole-exome sequencing and analysis

Exome sequencing was performed on DNA from skin fibroblasts andCD141 blood cells. Details on sample collection, extraction, library

preparation, capture, sequencing, and variant detection (IntegraGen, Evry,France) are provided in supplemental Methods. Indels and nonsynonymousexonic SNP with coverage >103, present in CD141 cells but not infibroblasts, were verified by bidirectional sequencing.

Liquid cell culture

Short-term culture of 5 3 104 cells and a single-cell culture of CD341

fractions were performed for 3 and 12 days, respectively, in minimumessential medium–alpha milieu with 10% fetal bovine serum (FBS) andrecombinant human cytokines: stem cell factor (SCF, 50 ng/mL), FLT3-ligand (50 ng/mL), pegylated thrombopoietin (TPO, 10 ng/mL), interleukin-3(IL-3, 10 ng/mL), interleukin-6 (IL-6, 10 ng/mL), granulocyte-macrophagecolony–stimulating factor (GM-CSF, 5 ng/mL), erythropoietin (EPO, 1 IU/mL), and granulocyte colony–stimulating factor (G-CSF, 10 ng/mL). Detailedprotocols are provided in supplemental Methods.

Methylcellulose colony–forming cell (CFC) assays

CD341 fractions were seeded in triplicate at 250 to 1000 cells per 1-mLculture dish in 2% standard methylcellulose supplemented with 37% FCS,12% bovine serum albumin, 1% l-glutamine, 1026 M of b-mercaptoethanol(Sigma-Aldrich, St. Louis, MO) and 20 ng/mL of SCF, 10 ng/mL of IL-3,3 IU/mL of EPO, and 10 ng/mL of G-CSF. Colonies were counted on day14. Clonogenic assays for TET2-mutated MDS and MPN are described insupplemental Methods.

Cell cycle analysis

CD341 BM cells were incubated with 10 mg/mL of Hoescht 33342 (LifeTechnologies Invitrogen, Grand Island, NY) for 1 hour at 37°C in minimumessential medium–alpha milieu with 10% FBS and cytokines (50 ng/mL ofSCF, 50 ng/mL of FLT3-ligand, 10 ng/mL of TPO, 10 ng/mL of IL-3, and10 ng/mL of IL-6). The cells were then stained, acquired, and analyzed asdescribed in supplemental Methods.

Serial replating assays

Clones generated at day 12 in liquid culture in the above-mentionedconditions were individually seeded in 96-well plates in 2% standardmethylcellulose with 37% FCS, 12% bovine serum albumin, 1% l-glutamine, 1026 M of b-mercaptoethanol and 20 ng/mL of SCF, 10 ng/mLof IL-3, and 10 ng/mL of G-CSF. At day 14, wells containing cell clusters(>20 cells) were counted, picked, washed, and reseeded in 96-wellmethylcellulose plates in similar conditions. This process was repeated 5times. Results are a percentage of the initial number of clones.

Genotyping of clones and colonies

DNA from liquid-culture clones or methylcellulose colonies was preparedas described previously.22 Mutational status was analyzed by mutation-specific fluorescent competitive probes with TaqMan real-time polymerasechain reaction (PCR) on an ABI 7500 (Applied Biosystems, Foster City,CA). Primers and probes for each mutation are detailed in supplementalMethods. Equivocal results were validated by sequencing.

Real-time quantitative PCR

Total RNA isolated with Trizol (Life Technologies) was reverse-transcribedwith Superscript Vilo (Life Technologies Invitrogen). Real-time PCR wasperformed on an Applied Biosystems 7500 Fast thermocycler using theSyBrGreen protocol (Applied Biosystems). Primers sequences will be givenon request.

TET2 knockdown by lentiviral delivery of shRNA

Sorted human cord bloodCD341/CD382 andCD341/CD381 cells were trans-duced with lentiviruses expressing the green fluorescent protein (GFP)and either shRNA-TET2 (59-GGGTAAGCCAAGAAAGAAA-39) orshRNA-scramble (59-GCCGGCAGCTAGCGACGCCAT-39) as described

BLOOD, 21 MARCH 2013 x VOLUME 121, NUMBER 12 CLONAL ARCHITECTURE IN CMML 2187

For personal use only.on May 30, 2017. by guest www.bloodjournal.orgFrom

previously.23 GFP1 CD341/CD382 or CD341/CD381 were sorted andplated on methylcellulose as described.

Statistical analyses

Statistical tests are indicated in corresponding legends. All tests were 2tailed. Statistical analyses were performed with StatView (SAS Institute,Cary, NC) or Prism (GraphPad Software, San Diego, CA).

Results

Clonal dominance in the early compartments of

hematopoietic differentiation

We first sorted peripheral blood CD141 cells of 28 patients withCMML and searched for mutations in 18 candidate genes encodingsignaling molecules (CBL, NRAS, KRAS, JAK2, FLT3, and KIT),epigenetic regulators (TET2, IDH1, IDH2, DNMT3A, ASXL1,EZH2, RUNX1, and NPM1), and splicing proteins (SRSF2, SF3B1,U2AF35, and ZRSR2). The number of mutations identified was 1,2, 3, and 4 in 9 patients (32%), 9 patients (32%), 7 patients(25%), and 3 patients (11%), respectively, including 2 patientswith 2 concomitant TET2 mutations. TET2, SRSF2, ASXL1, andsignaling genes (CBL, NRAS, KRAS, or JAK2) were mutated in 17patients (61%), 12 patients (43%), 7 patients (25%), and 12 patients(43%), respectively (Table 1). ASXL1 c.1934→p.G646WfsX12 wasverified to be a somatic mutation by the TaqMan allele dis-crimination assay, using skin fibroblasts as germline DNA (sup-plemental Figure 1).

Next, in 11 patients with available BM samples, we sortedCD341/CD382/CD901 (hematopoietic stem cells [HSC]), CD341

/CD382/CD902 (multipotent progenitors [MPP]), CD341/CD381/CD45RA2/CD1231 (common myeloid progenitors [CMP]), andCD341/CD381/CD45RA1/CD1231 (granulomonocyte progeni-tors [GMP]; supplemental Figure 2B) cells. Phenotypic profiles ofCD341 compartments were consistent between CMML and con-trols (supplemental Figure 2C), and their functional relevance inCMML was validated by testing their clonogenic potential (GMP,CMP) and their serial replating capacity (MPP, HSC; supplementalFigure 3G-H). In the 17 patients for whom only peripheral blood(PB) was available, CD341 cells were fractionated in CD341/CD382 and CD341/CD381 populations.

These subpopulations were cultured at 1 cell per well in liquidmedium for 12 days in the presence of a broad panel of cytokines(SCF, FLT3-L, TPO, IL-3, IL-6, G-CSF, GM-CSF, and EPO) tomaximize cloning efficacy and limit biases in clonal representation.We then collected the clones and performed TaqMan discrimina-tion analysis specifically designed for each individual mutationidentified initially in PB CD141 cells (supplemental Figure 2A).This strategy allows a faithful representation of clonal repartitionfor the following reasons: (1) The cloning efficacy of CD341

fractions was not influenced by the mutational status (supplementalFigure 3A-B). (2) Comparison of clones generated in theseconditions with colonies grown in methylcellulose showed anoverrepresentation of mutated clones in methylcellulose (supple-mental Figure 3C). (3) For 2 mutations, proportions of mutatedclones were similar when assessed by pyrosequencing or next-generation sequencing of bulk CD341 fractions (supplementalFigure 3D-F).

In all 28 patients, at least one of the mutations was found in morethan 75% of PB CD341/CD382 or BM HSC clones (Figure 1A),

without significant difference between CMML-1 and CMML-2cases. In 9 of 28 patients with sufficient materiel, 7 of 16 studiedmutations originally detected in peripheral blood CD141 cells werealso found in sorted CD31, including 2 of 7 in TET2, 2 of 3 inASXL1, 1 of 2 inCBL, 1 of 1 in JAK2, and 1 of 1 inKRAS but 0 of 2 inRUNX1 (examples in Figure 1B). Taken together, our results pointout an early amplification of the CMML clone at the CD341/CD382

stage of hematopoiesis.

Linear accumulation of mutations and clonal selection during

myeloid differentiation

We next analyzed the clonal hierarchy of CD341/CD382 cells orHSC of patients with 2 or more mutations. The genetic classi-fication of individual cells by allelic discrimination assays alloweda designation of subclones and the assembly of putative evolu-tionary trees. In most patients, mutations were found to accumulatein a linear succession, allowing the reconstitution of their sequentialacquisition. Detailed examples for 2 patients with 3 mutations arepresented in Figure 1C. Only 1 patient (unique patient number[UPN] #752) displayed somatic mosaicism (ie, acquisition of 2different mutations, NRAS G12D and KRAS G12A), in distinctsubclones (Figure 1D). The resulting succession of mutations inthe 28 patients is recapitulated in Table 1. In 4 cases, an orderof acquisition could not be assigned to 2 mutations (allinvolving a splice gene) because they were detected in thesame clones. Although the order acquisition of the mutationswas not fixed, mutations in TET2 or ASXL1 preceded mutationsin signaling genes in most cases.

We then studied the dynamics of clonal architecture duringmyeloid differentiation by comparing clonal frequencies in sortedHSC, MPP, CMP, and GMP. We found an overrepresentation ofsubclones with greater number of mutations in GMP comparedwith HSC/MPP (examples in Figure 1C). In 14 of 19 patients withmore than 1 mutation and sufficient numbers of assessable clones,the first event was present in more than 85% of HSC/MPP, withoutsignificant change in the frequency in CMP and GMP, under-scoring the early dominance of the leukemic clone, whereas thesecond event was significantly more frequent in GMP than in HSC/MPP (Figure 1E). Altogether, our results show linear acquisition ofmutations, clonal dominance of the malignant clone at the CD341/CD382 stages (HSC and MPP) of hematopoiesis, and furtherselection of the more mutated subclones during the early steps ofmyeloid differentiation until the GMP stage. This trend suggeststhat secondary mutations may provide an advantage to the cloneduring myeloid differentiation.

Clonal branching through loss of heterozygosity

We next took advantage of information on heterozygosity andhomozygosity of mutations to refine the evolutionary trees of the19 patients with 2 or more mutations. In 10 patients, the archi-tecture was strictly linear and mosaicism was identified in 1 pa-tient. In the remaining 8 patients, we identified subclones with lossof heterozygosity (LOH) of mutations in CBL (n 5 3, including 2with a concomitant LOH in the TET2 or SF3B1 loci, respectively),KRAS, JAK2, TET2, and ZRSR2 (each n 5 1), giving rise to singleor multiple branching in the clonal architecture (example inFigure 2A, summarized in Table 1). The detection of subclones withreversion of a mutation to wild-type status suggested a mitoticrecombination event as the origin of LOH. This hypothesis wasassessed by taking advantage of an informative intronic singlenucleotide polymorphism (SNP, rs77741206G/A) located 46 base

2188 ITZYKSON et al BLOOD, 21 MARCH 2013 x VOLUME 121, NUMBER 12

For personal use only.on May 30, 2017. by guest www.bloodjournal.orgFrom

Table

1.Clinicalandmolecularcharacteristicsofthe28patients

studiedattheclonallevel

UPN

Age

Sex

(M/F)

WHO

category

Karyotype

Monocytes(g/L)

No.ofmutations

Firstmutation

Secondmutation

Thirdmutation

Fourthmutation

Architecture

428

74

FCMML-1

Norm

al

5.1

1TET2(M

993VfsX15)

-

462

83

MCMML-2

Norm

al

5.0

2ASXL1(E635RfsX15)

RUNX1(G

60PfsX52)

Linear

495

53

MCMML-1

Norm

al

1.9

2CBL(S407P)

SRSF2(P95H)

Branched(LOH

CBL)

497

59

MCMML-1

Norm

al

2.1

1TET2(L757TfsX12)

-

498

84

MCMML-1

Trisomy13

2.2

3ASXL1(G

646WfsX12)*

ZRSR2*

(splicen11ex9)

SRSF2(P95L)

Branched(LOH

ZRSR2)

500

86

FCMML-2

Norm

al

3.4

1TET2(P1619LfsX4)

-

501

74

FCMML-1

Norm

al

1.1

1TET2(Q

8033)

-

507

70

MCMML-2

Norm

al

4.9

3SRSF2(P95H)

CBL(W

408R)

TET2(S7163)

Branched(LOH

CBL&TET2)

514

80

FCMML-1

Norm

al

1.8

1KRAS(G

12S)

-

516

76

MCMML-1

Norm

al

7.4

2TET2(K4503)

NRAS(G

12D)

Linear

518

51

MCMML-1

Norm

al

1.6

1KRAS(G

12C)

-

524

90

MCMML-1

Norm

al

38.9

3TET2(Q

17023)

SRSF2(P95L)

TET2(S1525YfsX53)†

KRAS(A18D)

Linear

531

81

FCMML-1

Norm

al

3.4

3TET2(Q

8103)*

SRSF2(P95L)*

JAK2(V617F)

Linear

536

79

FCMML-1

Norm

al

1.4

2CBL(R

420L)

TET2(S16303)

Branched(LOH

CBL)

537

57

MCMML-1

Norm

al

3.2

2TET2(E8523)

SRSF2(P95H)

Linear

550

85

MCMML-1

Norm

al

2.7

3TET2(D

1242TfsX11)

SRSF2(P95H)

RUNX1(Y376LfsX197)

Branched(LOH

TET2)

586

61

FCMML-1

ND

10.3

2ASXL1(G

646WfsX12)

TET2(N

275IfsX18)

Linear

632

64

MCMML-2

Norm

al

10.0

3TET2(N

1156I)

SRSF2(P95H)

NRAS(A59G)

Linear

638

80

MCMML-1

Norm

al

2.6

4SF3B1(D

586H)

CBL(C

404Y)

SRSF2(P95R)

ASXL1(G

6423)

Branched(LOH

CBL&SF3B1)

644

82

FCMML-1

Norm

al

1.1

4ASXL1(G

646WfsX12)*

SRSF2(P95H)*

TET2(S1190YfsX3)

TET2(R

1214Q)

Linear

654

70

FCMML-1

Trisomy8

2.3

3ASXL1(G

646WfsX12)

SRSF2(P95L)

JAK2(V617F)

Linear

658

67

MCMML-2

Monosomy7

2.62

1ASXL1(G

646WfsX12)

-

662

89

MCMML-1

ND

3.37

1TET2(V1454SfsX4)

-

731

78

MCMML-1

ND

11.0

2SRSF2(P95L)

JAK2(V617F)

Branched(LOH

JAK2)

736

73

MCMML-1

Norm

al

5.1

1KRAS(T58I)

-

743

82

FCMML-1

Norm

al

2.8

2TET2(Q

3233)

TET2(Q

796TfsX6)

Linear

752

76

MCMML-2

del5q1

tri8

17.3

2NRAS(G

12D)‡

KRAS(G

12A)‡

Branched(m

osaicism)

759

81

MCMML-1

ND

1.5

3TET2(T1047SfsX9)*

U2AF35(Q

157R)*

KRAS(K117R)

Branched(LOH

NRAS)

*exaequomutations.

†2mutationsin

TET2.

‡Clonalmosaicism.

BLOOD, 21 MARCH 2013 x VOLUME 121, NUMBER 12 CLONAL ARCHITECTURE IN CMML 2189

For personal use only.on May 30, 2017. by guest www.bloodjournal.orgFrom

Figure 1. Early clonal dominance, linear accumulation of mutations, and clonal selection during myeloid differentiation. (A) Clones from sorted BM HSC (CD341/

CD382/CD901, noted with an asterisk) or peripheral blood immature cells (CD341/CD382) of 28 patients (detailed in Table 1) were cultured at 1 per well in liquid medium for

12 days in the presence of cytokines, then collected, and assessed by allele discrimination analysis of each individual mutation initially identified in peripheral blood CD141

cells. Proportion of clones with black: wild type, yellow: 1 mutation, orange: 2 mutations, red: 3 mutations, purple: 4 mutations. Numbers on the top of the bars indicate the

number of clones analyzed. (B) Examples of Sanger sequences in CD141 and CD31 sorted cells for 2 mutations in a representative sample (UPN #524). (C) Putative

evolutionary trees of sorted BM CD341 populations generated by genetic classification of individual cells. Simple linear trees from 2 samples (top: UPN #531: TET2 Q8103;

SRSF2 P95H; JAK2 V617F; bottom: UPN #632: TET2 N1156I; SRSF2 P95L; NRAS A59G) with all mutations heterozygous in all clones (white: no mutation, yellow: 1

mutation, orange: 2 mutations, red: 3 mutations). (D) Repartition of 193 CD341/CD382 clones showing a unique somatic mosaicism in patient UPN #752 harboring 2

2190 ITZYKSON et al BLOOD, 21 MARCH 2013 x VOLUME 121, NUMBER 12

For personal use only.on May 30, 2017. by guest www.bloodjournal.orgFrom

pairs 39 from SF3B1 c.1756G.C (p.D586H) in 1 patient (UPN#638). The SNP was heterozygous in 9 clones with a heterozy-gous SF3B1 mutation, whereas symmetric LOH of rs77741206was detected in 9 clones with a wild-type or homozygous SF3B1mutation (Figure 2B). The clonal diversity generated by thesebranched architectures was also submitted to clonal selectionduring myeloid differentiation to the CMP and GMP stages(example in Figure 2C).

To confirm that the clonal architecture identified by focusing on18 frequently mutated genes holds true when analyzing the fullspectrum of genetic lesions, we performed whole-exome sequencingof peripheral CD141 cells in a patient with TET2 and U2AF35mutations. We identified 14 additional gene mutations confirmedby Sanger sequencing, including a noncanonical KRAS mutation(K117R). All mutations were heterozygous except the homozygousTET2 deletion. Again, mutation-specific discrimination analysis ofCD341 colonies after liquid culture showed the linear accumu-lation of mutations, with only minor subclones possibly resultingfrom a mitotic recombination affecting the KRAS locus (supplementalFigure 4). Altogether, the dominant clone mostly results fromsequential waves of mutation acquisition and expansion, with minorsubclones probably generated by homologous recombination atspecific sites.

Clonal architecture is dynamic and is selectively affected

by therapy

We analyzed the fate of clonal heterogeneity in the face of temporalevolution of CMML, either in untreated patients or in patients withcurrently available therapies (detailed in supplemental Table 1). Inan untreated and clinically stable patient analyzed twice at a 12-month interval, the proportion of clones with 2 mutations increasedsignificantly in HSC/MPP (UPN #507; Figure 3A). A similarevolution was noted, with the same interval in a patient receivinghydroxyurea (UPN #516; Figure 3B). In a patient with stable diseaseunder hypomethylating therapy (UPN #550), we observed theamplification of the more mutated clone, which harbors an ASXL1mutation (Figure 3C). It is surprising to note that when UPN #506achieved erythroid improvement with an erythropoiesis-stimulatingagent (after natural evolution depicted in Figure 3B), a re-expansionof wild-type hematopoiesis was observed in the CD341/CD382

compartment, although the selective advantage of the more mutatedclone persisted during myeloid differentiation (Figure 3D). In UPN#752 with a somatic mosaicism betweenKRAS and NRASmutations,achievement of a transient control of myeloproliferation and cir-culating blast counts with an investigational mitogen-activated proteinkinase kinase inhibitor only resulted in a shift of the equilibriumbetween the 2 clones, with a slight expansion of the KRAS-mutatedclone at the expanse of the NRAS mutation (supplemental Figure 5).Finally, 1 patient (UPN #632) with proliferative CMML achievedpartial response with intensive chemotherapy, converted to completeresponse after hypomethylating therapy, then underwent allogeneicstem cell transplantation and experienced a relapse 6 months later.Sequential analysis of his clonal architecture showed a reduction of theNRAS-mutated clone by chemotherapy and hypomethylating therapy,allowing the expansion of the ancestral TET2/SRSF2-mutated clone.Nonetheless, the more mutated clone persisted and regained domi-nance at relapse (Figure 3E).

GMP are quantitatively and qualitatively normal in CMML

To understand the relationship between the clonal architecture andthe phenotype of CMML, we sought to analyze the cellularmechanisms driving the emergence from the CD341 compartmentof the granulomonocytic expansion of CMML. Because clonaldominance, notably regarding secondary mutations, is achieved atthe GMP stage (Figure 1E), we quantitatively and qualitativelyanalyzed this compartment. The differentiation potential of GMPwas assessedwith the gold-standard 14-daymethylcellulose assay, inthe presence of FBS and a restricted panel of cytokines (SCF, IL-3,G-CSF, and EPO) compared with previous experiments (Figure 1-3).This panel allowed comparison of the granulomonocytic andpotential for erythroid differentiation of all CD341 fractions, andavoided the use of GM-CSF whose activity is heterogeneous inCMML.24-26 The potential for self-renewal was assessed by serialreplating in similar conditions without EPO to focus on granulomo-nocytic progenitors. By flow cytometry, the percentage of GMPamong BMCD341 cells was not significantly different in 33 patientswith CMML (with a mutation spectrum representative of the entirepopulation, not shown) and 15 age-matched healthy participants(Figure 4A). The ability to form granulomonocytic colonies inmethylcellulose was similar in patients and healthy controlparticipants (Figure 4B). Cell cycle analysis did not identify anysignificant change in patient cells (Figure 4C), and serial replating didnot demonstrate a strong increase in their self-renewal capability(Figure 4D). Overall, no systematic quantitative or qualitativealteration of GMP was identified in CMML, suggesting that thegranulomonocytic expansion characteristic of this disease may haveappeared at earlier stages of myeloid differentiation.

CMP and MPP undergo premature granulomonocytic

differentiation in CMML

Contrasting with the normal GMP fraction, the percentages ofCMP and MPP were higher in patients with CMML than in healthycontrol participants (Figure 5A). When cultured in methylcellulosein similar conditions as GMP, both CMP and, to a lesser extent,MPP demonstrated an increased ability to form granulomonocyticcolonies at the expense of erythroid colonies (Figure 5B). Similarresults were found when CMP were cultured as single cells ina liquid medium for 12 days in the presence of a larger panel ofcytokines (SCF, FLT3-L, TPO, IL-3, IL-6, G-CSF, GM-CSF, andEPO), with a strong increase in the percentage of granulomono-cytes (CD141 and/or CD151 and/or CD241) at the expense oferythroid (GPA1) clones (supplemental Figure 6B), despiteunchanged cloning efficacy (supplemental Figure 6A). Analysisof mutations in granulomonocytic and erythroid colonies obtainedby culturing CMP from 3 patients showed a similar repartition ofthe mutated alleles in the 2 types of colonies, demonstrating thatthe majority of erythroid colonies belong to the most mutatedsubclone (not shown). Short-term (3-day) culture of CMP in liquidmedium followed by immunophenotypic analyses showed theirincreased ability to mature into GMP at the expense of MEP(Figure 5C). The functional relevance of CD341 cell immunophe-notyping after this 3-day culture was validated by resorting GMP,MEP, and CMP at day 3 and assessing their clonogenic potential(supplemental Figure 6C). Gene expression analysis revealed an

Figure 1 (continued) independent subclones with NRAS G12D and KRAS G12A genotypes. (E) Proportion of mutated clones in HSC or MPP, CMP, and GMP fractions from

14 patients (10 with 2 or more mutations, 4 with 1 mutation); ns, nonsignificant, *P , .05 (HSC or MPP vs GMP, Wilcoxon matched-pair signed-rank test for 11 and 8 pairs,

respectively). Bar: median percentage.

BLOOD, 21 MARCH 2013 x VOLUME 121, NUMBER 12 CLONAL ARCHITECTURE IN CMML 2191

For personal use only.on May 30, 2017. by guest www.bloodjournal.orgFrom

Figure 2. Clonal branching through loss of heterozygosity by mitotic recombination. (A) Putative evolutionary tree generated from the classification of all sorted BM

CD341 clones (n5 106; pooling all subpopulations) for UPN #638 accounting for the heterozygosity status at each locus, showing multiple branching with LOH of the CBL and

SF3B1 mutation (italicized: heterozygous, bold & underlined: homozygous). (B) Corresponding Sanger sequences at the SF3B1 locus of CD341 clones from UPN #638 with

wild-type, heterozygous, and homozygous SF3B1 according to the TaqMan allelic discrimination assay showing the mutation region (top panel) and the informative SNP

rs77741206 located 46 bp in the 39 intronic region. Sequences from total CD141 cells of UPN #638 and from the UT7 cell line as control. (C) Putative evolutionary trees of

sorted BM CD341 populations from UPN #507. White: no mutation, yellow: 1 mutation, orange: 2 mutations, red: 3 mutations, purple: 4 mutations. Only mutated genes are

indicated in each subclone (italicized: heterozygous, bold and underlined: homozygous).

2192 ITZYKSON et al BLOOD, 21 MARCH 2013 x VOLUME 121, NUMBER 12

For personal use only.on May 30, 2017. by guest www.bloodjournal.orgFrom

increased expression of PU.1 with nonsignificant changes in thelevel of GATA1, CEBPA, and CEBPB in CMP sorted from CMMLcompared with those from healthy control participants (Figure 5D;supplemental Figure 6D). Finally, the proportion of granulomo-nocytic colonies differentiating from CMP at the expense oferythroid colonies correlated inversely with patient hemoglobinlevel (Figure 5E). Altogether, these results suggest that immaturepluripotent progenitors (CMP and, to a lesser extent, MPP) areskewed toward granulomonocytic differentiation in CMML.

Early clonal dominance of TET2 mutations leads to

granulomonocytic hyperplasia

To unravel a link between early clonal dominance and myeloiddifferentiation skewing in CMML, we focused on TET2, which is themost frequently6,7,27 and early (Table 1) mutated gene in CMMLknown to date. TET2 is also mutated in 15% to 25% of cases of MDSand MPN,5,27,28 where granulomonocytic hyperplasia is absent(MDS) or variable (MPN).We interrogated the stage of expansion ofthe TET2-mutated clone in colonies generated from CD341/CD382

and CD341/CD381 cells collected from TET2-mutated MPN (n 58) orMDS (n5 5) (supplemental Table 2).We observed a significantcorrelation between the proportion of cells with mutated TET2 at theCD341/CD382 level and peripheral monocyte count (Figure 6A, leftpanel). Such a correlation was not found when the more matureCD341/CD381 compartment was considered (Figure 6A, rightpanel). In 5 patients with CMML, the ability of CMP to form

granulomonocytic colonies reflected the size of the TET2-mutatedclone in this compartment (Figure 6B). Finally, we recapitulatedTET2 invalidation by lentiviral transduction of a previously reportedshRNA directed against TET223 in sorted CD341/CD382 andCD341/CD381 normal progenitors, followed by culture in thepresence of erythroid and granulomonocytic cytokines. Functionalknockdown of TET2 in CD341/CD382 caused a granulomonocyticexpansion that was not observed in CD341/CD381 cells(Figure 6C). Altogether, early dominance of the TET2-mutatedclone in the immature CD341/CD382 compartment mayparticipate in the granulomonocytic skewing that defines CMML.

Discussion

CMML shares with MPN, MDS, and other MDS/MPN severalgene mutations. Whereas a driver oncogenic mutation leading tothe constitutive activation of intracellular signaling pathways hasbeen identified in most MPN16 and in juvenile myelomonocyticleukemia,29 the driver oncogenic events involved in CMML areless clearly identified. TET2, SRSF2 (which is often associated toTET2),6,13 and ASXL1 are the most frequent genetic eventsin CMML.6,7,14,15 The number of mutated genes is a strongprognostic factor in CMML.7 This prompted the investigation ofthe clonal architecture of genetic lesions in 28 cases of CMML,

Figure 3. Clonal architecture is dynamic and selectively affected by therapy. Proportion of mutated clones in the indicated peripheral blood (A) or BM (B-E) CD341

fractions of matched samples from untreated (A) or treated (B-E) patients. Mutated genes are indicated in their inferred order of apparition from top to bottom (see Table

1 for details). (A) UPN #516: TET2 K4503, NRAS G12D. (B,D) UPN #507: TET2 S7163, CBL W408R, SRSF2 P95H. C. UPN #550: TET2 D1242TfsX11, RUNX1

Y376LfsX197, SRSF2 P95H, ASXL1 E635RfsX15.(E) UPN #632: TET2 N1156I; NRAS A59G; SRSF2 P95H. The total number of interrogated clones is indicated on the

top of the bars. Missing fractions represent phenotypes underrepresented in posttreatment samples. Note that in UPN #632, showing RAEB-2 rapidly evolving to AML at relapse,

the LMPP and GMP fractions were dominant, as previously described.47 12 mo, 12 months untreated evolution; ASCT, Allogeneic Stem Cell Transplantation; ESA, Erythropoiesis-

Stimulating Agent; HMA, Hypomethylating Agent; HY, Hydroxyurea; IC, Intensive Chemotherapy; LMPP, Lymphoid-primed multipotent progenitors; Re, AML Relapse.

BLOOD, 21 MARCH 2013 x VOLUME 121, NUMBER 12 CLONAL ARCHITECTURE IN CMML 2193

For personal use only.on May 30, 2017. by guest www.bloodjournal.orgFrom

showing that gene mutations are sequentially acquired, leading toa mostly linear clonal architecture, with frequent branching throughLOH resulting frommitotic recombinations. Early clonal dominance atthe CD341/CD382 stage of hematopoiesis, especially dominance of

TET2-mutated cells, is a key feature of the disease that is accompaniedby myeloid differentiation skewing toward the granulomonocyticlineages, whereas additional genetic events rapidly confer a selectiveadvantage to the clone during myeloid differentiation.

Figure 4. GMP are quantitatively and qualitatively

normal in CMML. (A) Proportion of GMP (CD341/

CD381/CD45RA1/CD1231) by flow cytometry among

total BM CD341 cells from CMML or aged-matched

healthy control participants. Bar: median. (B) Total

number of colonies (all granulomonocytic) per 1000

GMP grown in triplicate in methylcellulose for 14 days

with 30% FBS and cytokines (SCF, IL-3, G-CSF, and

EPO) from 5 control participants and 16 patients with

CMML. (C) Fractions of GMP cells from fresh BM

samples in G0/G1 (white), S (gray), and G2/M (black)

phase after 1-hour incubation in Hoescht 33342 in the

presence of 10% FBS and cytokines; mean and SD from

independent samples. (D) Serial replating in methylcel-

lulose of individual clones from sorted GMP of patients

with CMML (solid) or control participants (dashed).

Results: percentage of colonies relative to the initial

number of clones seeded. Mean and SD from 3

independent samples for each group (*P , .05, Mann-

Whitney U test); ns, nonsignificant.

Figure 5. CMP and MPP undergo premature granulomonocytic differentiation in CMML. (A) Proportion of MPP (CD341/CD382/CD45RA2/CD902, top panel) and CMP

(bottom panel, CD341/CD381/CD45RA2/CD1231) by flow cytometry among total BM CD341 cells from patients with CMML (n 5 33) or aged-matched healthy control

participants (n 5 15). Bar: median; Mann-Whitney U test. (B) Proportion of mixed (CFU-M, white), granulomonocytic (CFU-G/M: total of CFU-G, CFU-M, and CFU-GM; black)

and erythroid colonies (BFU-E, gray) colonies from 250 MPP (top) or CMP (bottom) grown in triplicate in methylcellulose for 14 days with 30% FBS and cytokines (SCF, IL-3,

G-CSF, and EPO); x2 test for mean proportions from CMML (CMP: n 5 18; MPP: n 5 9) and control (CMP: n 5 5; MPP: n 5 3) samples. (C) 5 3 104 CMP from healthy (n 5

4) or CMML (n 5 8) samples were cultivated in bulk for 3 days in the presence of 10% FBS and cytokines (SCF, FLT3-L, TPO, IL-3, IL-6, G-CSF, GM-CSF, and EPO) and

then were analyzed by flow cytometry for CD34, CD38, CD45RA, and CD123 expression. Representation of the ratio of events with a GMP (CD341/CD381/CD45RA1/

CD1231) phenotype to cells with an MEP CD341/CD381/CD45RA2/CD1232) phenotype; bar: median. (D) Gene expression levels (relative to HPRT expression) of PU.1 (top

panel) and GATA1 (bottom) in sorted CMP fractions: mean and SEM from 4 control and 5 CMML samples each analyzed in duplicate. Unpaired t tests. *P , .05; **P , .01;

***P , .001. (E) Correlation between the percentage of granulomonocytic colonies from CMP of 16 CMML samples and the corresponding patient’s hemoglobin level (g/dL).

Spearman test, slope from linear regression.

2194 ITZYKSON et al BLOOD, 21 MARCH 2013 x VOLUME 121, NUMBER 12

For personal use only.on May 30, 2017. by guest www.bloodjournal.orgFrom

Tumor genetic heterogeneity was initially thought to evolvethrough sequential, “linear” acquisition of collaborating mutations,followed by waves of clonal expansion.30 Recent analyses in acuteleukemias revealed more complex, oligoclonal models.2,4,31 Here, weshow that the clonal architecture of CMML may reconcile bothmodels (ie, a mostly linear acquisition of the mutations), with a fewramifications likely the result of homologous recombinations.9 LOHoccurs at loci-encoding genes from all classes (epigenetic, splicing,and signaling), though homozygous clones achieving clonaldominance are mostly restricted to the TET2 and CBL genes.5,9

Next-generation sequencing identifies an average of 10 somaticexonic mutations in MDS, all of which arise in the same CD341

cell.4,32 In all but one of the 28 studied cases of CMML, all mutationswere present in at least one CD341/CD382(/CD901). Exomesequencing in CD341 leukemic clones from a representative patientwith CMML followed by clonal architecture analysis identifiedsequential waves of acquisition of mutations, each wave including 1presumed driver oncogenic lesion, such as TET2, U2AF35, or KRASmutations. Sequential analyses demonstrated that, aside fromallogeneicstem cell transplantation, all current therapeutic strategies modu-late, rather than eradicate, the leukemic clone.

The reconstructed phylogenetic trees also allowed ordering ofthe sequence of acquisition of the studied mutations, except in 4cases, raising the possibility that 2 mutations were concomitantly

gained in those cases. Why certain mutations, such as those inTET2 and SRSF2, are often associated remains uncertain, but thefinding that their order of acquisition is biased but not stereotypedsuggests that their association arises from functional cooperation.33

Clonal dominance is a heterogeneous feature in myeloidmalignancies. For example, clonal expansion of JAK2V617Fmutationsoccurs in earlier progenitors in myelofibrosis34 than in polycythemiavera,22 a difference attributed to the role of additional mutations inTET25 or ASXL1.19 Here, we show that early clonal dominance ofinitial mutations in CD341/CD382 immature progenitors is a keyfeature of CMML. Clones harboring only part of the oncogeniclesions persist in this compartment, whereas the fully mutatedleukemic clone harboring secondary lesions is further selected duringmyeloid differentiation to the GMP stage. Mutations affecting cellsignaling such as JAK2, NRAS, or KRAS mutations, or LOH of aheterozygousCBLmutation, may favor this differentiation-associatedselection, as JAK2V617F does in MPN.22

Our strategy to establish clonal architecture, based on the trackingof recurrent mutations in single-cell–derived colonies after in vitroculture, had 2 limitations. First, it did not address the full spectrum ofgenetic aberrations present in the tumor. Founder mutations couldhave preceded what we detect as “primary” events. If this is the case,early clonal dominance would be of even greater magnitude thanreported. In the unique patient studied by whole exome, however, the

Figure 6. Early dominance of TET2 mutations contributes to granulomonocytic skewing in CMML. (A) Correlation between the proportion of TET2-mutated clones of

CD341/CD382 cells grown in liquid culture with cytokines with (n 5 11) or without (n 5 2) a stromal layer (top panel), or corresponding CD341/CD381 cells grown in

methylcellulose (n 5 11) or in liquid culture (n 5 2; bottom panel) from 13 TET2-mutated samples (8 MPN, 5 MDS). Spearman test and slope from linear regression. (B)

Percentage of CMP clones with mutated TET2 assessed in liquid culture (gray bars) compared with the percentage of granulomonocytic colonies after methylcellulose culture

of CMP in the presence of SCF, IL-3, G-CSF, and EPO (black bars). Results are from 5 samples. Numbers on top of the gray bars indicate the number of CMP clones

assessed. UPN: unique patient number. (C) Sorted CD341/CD382 and CD341/CD381 cord blood cells were transduced with an shRNA directed against TET2 or a scrambled

shRNA (as previously described23). After infection, GFP1 cells were cultured as in (B). At day 14, cells were washed and stained with PE-anti-GPA and APC-anti-CD33.

Ratios of GFP1 granulomonocytic (CD331/GPA2) to erythroid (CD336/GPA1) cells from 4 independent experiments are represented (mean 6 SEM, Mann-Whitney U test,

*P , .05).

BLOOD, 21 MARCH 2013 x VOLUME 121, NUMBER 12 CLONAL ARCHITECTURE IN CMML 2195

For personal use only.on May 30, 2017. by guest www.bloodjournal.orgFrom

TET2 “primary” event was present in the founder clone, in accordancewith the finding of TET2 mutations in preleukemic HSC in AML35

and in HSC of older patients with clonal, but nonmalignant,hematopoiesis.36 Second, in vitro culture may introduce biases inclonal representation. However, validation of single-cell–derivedcolonies genotyping has been reported in AML35 and was confirmedin our present study. Next-generation sequencing represents themajoralternative strategy to assess clonal architecture, but it is either appliedto bulk samples4 with the risk of missing mosaic subclones, or toamplified DNA from single cells, thus introducing other biases andlimiting resolution with current techniques.37,38

Because genes mutated in CMML are also found in other myeloidmalignancies, we next sought to understand the relationship betweenthe clonal architecture of CMML and the emergence of its distinctivephenotypic trait, namely granulomonocytic hyperplasia. Whereasanalysis of sorted progenitors did not identify alterations of theGMP compartment, the CMML BM CD341 compartment wasenriched in CMP and MPP with a skewed myeloid differentiation(ie, an enhanced capability to form granulomonocytic colonies atthe expense of erythroid colonies). This skewed myeloiddifferentiation could result from biased lineage decision and/ordifferential amplification of granulomonocytic and erythroidprogenitors after lineage commitment, a question recentlyaddressed in normal hematopoietic differentiation by single-cell–tracking studies.39,40

Inhibition of TET2 expression in CD341 cells has been shown topromote granulomonocytic differentiation in vitro and in vivo,23,41,42

yet TET2 mutation allele burden in total BM–nucleated cells(representing differentiated cells) seems comparable in MDS, whichlacks monocytosis, and CMML.27 To elucidate this paradox, weanalyzed the amplification of TET2 mutations in CD341 progenitorcells from patients with either MDS, or those with JAK2V617F MPN,where monocyte count is variable.43 In these diseases, we uncovereda specific correlation between amplification of TET2-mutated cells inthe immature CD341/CD382 compartment and higher monocytecounts. This finding could be explained by a role of TET2 in restrict-ing a default granulomonocytic program in immature multipotent

progenitors. Such a hypothesis is reinforced by the observation thatfunctional invalidation of TET2 in healthy CD341/CD382, but not inCD341/CD381, cells, leads to granulomonocytic amplification.

We propose a model linking early clonal dominance withgranulomonocytic skewing in CMML (Figure 7). According to thismodel, early dominance of the mutated clone, in particular, in thepresence of a TET2 mutation, accounts for the accumulation ofmonocytes in CMML. In other chronic myeloid malignancies suchas MDS or MPN, TET2 mutations may reach clonal dominance atlater stages of myeloid differentiation (eg, because of additionalmutations such as JAK2V617F). Differences in the dynamics of clonalamplification of TET2mutations could be explained by either specificcombinations of mutations (eg, TET2/SRSF2 in CMML),6,13 theiroccurrence in distinct subsets of HSC,44 or a stochastic effect of TET2in the initially mutated cell (Figure 7). These models could explainwhy monocytosis can develop in the course of MDS45 or MPN46

without acquisition of novel mutations. In addition to providinghypotheses to explain the disease phenotype, analyses of clonalarchitecture may allow defining therapeutic strategies thatimprove disease control and possibly achieve clonal eradicationby targeting the first genetic aberrations that accumulate inimmature progenitors.

Acknowledgments

The authors thank Fatiha Chermat (GFM administrative officer)and Vladimir Lazar (genomic platform at Institut Gustave Roussy)for their assistance.

This work was supported by grants from the Ligue NationaleContre le Cancer (Label, E.S.), Agence Nationale de la Recherche(E.S.), Institut National du Cancer (PHRC 2011 to E.S. and supportto R.I.), Association Laurette Fugain (N.D.), and Fondation deFrance (N.D.). The UMR1009, IFR54, and GRC Saint-Antoineequipment was supported by the Association pour la Recherche surle Cancer and Region Ile-de-France.

Figure 7. Proposed model linking early clonal

dominance with granulomonocytic skewing of

immature progenitors in CMML. According to their

stage of expansion (early in CMML, later in other

chronic myeloid neoplasms), TET2 mutations will or will

not give rise to monocytosis. TET2-mutated cells are

represented with dashed lines. Granulomonocytic dif-

ferentiation potential is depicted in green, erythroid

potential in red.

2196 ITZYKSON et al BLOOD, 21 MARCH 2013 x VOLUME 121, NUMBER 12

For personal use only.on May 30, 2017. by guest www.bloodjournal.orgFrom

Authorship

Contributions: R.I. performed experiments and statistical analyses,analyzed data, and drafted the manuscript. O.K. and A.R. performedgenotyping and analyzed data. M.M. performed sample collection,qualification, and nucleic acid extractions. P.R. performed cellsorting. C.O. and G.M. performed and analyzed microarray data.F.D. provided samples and lentiviral vectors. C.B., L.A., P.F., U.P.,

and O.G. provided samples. C.P., O.A.B., M.F., W.V., and N.D.supervised genotyping and analyzed data. E.S. provided samples,analyzed data, supervised the work, and revised the manuscript. Allauthors approved the final manuscript.

Conflict-of-Interest disclosure: The authors declare no com-peting financial interests.

Correspondence: Eric Solary, Inserm UMR 1009, InstitutGustave Roussy, 114, Rue Edouard Vaillant, 94805 Villejuif,France; e-mail: eric.solary@igr.fr.

References

1. Greaves M, Maley CC. Clonal evolution in cancer.Nature. 2012;481(7381):306-313.

2. Anderson K, Lutz C, van Delft FW, et al. Geneticvariegation of clonal architecture and propagatingcells in leukaemia. Nature. 2011;469(7330):356-361.

3. Stephens PJ, Greenman CD, Fu B, et al. Massivegenomic rearrangement acquired in a singlecatastrophic event during cancer development.Cell. 2011;144(1):27-40.

4. Walter MJ, Shen D, Ding L, et al. Clonalarchitecture of secondary acute myeloid leukemia.N Engl J Med. 2012;366(12):1090-1098.

5. Delhommeau F, Dupont S, Della Valle V, et al.Mutation in TET2 in myeloid cancers. N Engl JMed. 2009;360(22):2289-2301.

6. Meggendorfer M, Roller A, Haferlach T, et al.SRSF2 mutations in 275 cases with chronicmyelomonocytic leukemia (CMML). Blood. 2012;120(15):3080-3088.

7. Jankowska AM, Makishima H, Tiu RV, et al.Mutational spectrum analysis of chronicmyelomonocytic leukemia includes genesassociated with epigenetic regulation: UTX,EZH2, and DNMT3A. Blood. 2011;118(14):3932-3941.

8. Bacher U, Haferlach T, Schnittger S, et al. Recentadvances in diagnosis, molecular pathology andtherapy of chronic myelomonocytic leukaemia. BrJ Haematol. 2011;153(2):149-167.

9. Dunbar AJ, Gondek LP, O’Keefe CL, et al. 250Ksingle nucleotide polymorphism array karyotypingidentifies acquired uniparental disomy andhomozygous mutations, including novel missensesubstitutions of c-Cbl, in myeloid malignancies.Cancer Res. 2008;68(24):10349-10357.

10. Levine RL, Loriaux M, Huntly BJ, et al. TheJAK2V617F activating mutation occurs in chronicmyelomonocytic leukemia and acute myeloidleukemia, but not in acute lymphoblastic leukemiaor chronic lymphocytic leukemia. Blood. 2005;106(10):3377-3379.

11. Tyner JW, Erickson H, Deininger MW, et al. High-throughput sequencing screen reveals novel,transforming RAS mutations in myeloid leukemiapatients. Blood. 2009;113(8):1749-1755.

12. Yoshida K, Sanada M, Shiraishi Y, et al. Frequentpathway mutations of splicing machinery inmyelodysplasia. Nature. 2011;478(7367):64-69.

13. Kar SA, Jankowska A, Makishima H, et al.Spliceosomal gene mutations are frequent eventsin the diverse mutational spectrum of chronicmyelomonocytic leukemia but largely absent injuvenile myelomonocytic leukemia.Haematologica. 2013;98(1):107-113.

14. Gelsi-Boyer V, Trouplin V, Adelaıde J, et al.Mutations of polycomb-associated gene ASXL1 inmyelodysplastic syndromes and chronicmyelomonocytic leukaemia. Br J Haematol. 2009;145(6):788-800.

15. Abdel-Wahab O, Pardanani A, Patel J, et al.Concomitant analysis of EZH2 and ASXL1mutations in myelofibrosis, chronic

myelomonocytic leukemia and blast-phasemyeloproliferative neoplasms. Leukemia. 2011;25(7):1200-1202.

16. Vainchenker W, Delhommeau F, ConstantinescuSN, et al. New mutations and pathogenesis ofmyeloproliferative neoplasms. Blood. 2011;118(7):1723-1735.

17. Bejar R, Stevenson K, Abdel-Wahab O, et al.Clinical effect of point mutations inmyelodysplastic syndromes. N Engl J Med. 2011;364(26):2496-2506.

18. Beer PA, Delhommeau F, LeCouedic JP, et al.Two routes to leukemic transformation aftera JAK2 mutation-positive myeloproliferativeneoplasm. Blood. 2010;115(14):2891-2900.

19. Stein BL, Williams DM, O’Keefe C, et al.Disruption of the ASXL1 gene is frequent inprimary, post-essential thrombocytosis and post-polycythemia vera myelofibrosis, but not essentialthrombocytosis or polycythemia vera: analysis ofmolecular genetics and clinical phenotypes.Haematologica. 2011;96(10):1462-1469.

20. Papaemmanuil E, Cazzola M, Boultwood J, et al.Chronic Myeloid Disorders Working Group of theInternational Cancer Genome Consortium. SomaticSF3B1 mutation in myelodysplasia with ringsideroblasts.NEngl JMed. 2011;365(15):1384-1395.

21. Braun T, Itzykson R, Renneville A, et al; GroupeFrancophone des Myelodysplasies. Molecularpredictors of response to decitabine in advancedchronic myelomonocytic leukemia: a phase 2 trial.Blood. 2011;118(14):3824-3831.

22. Dupont S, Masse A, James C, et al. The JAK2617V.F mutation triggers erythropoietinhypersensitivity and terminal erythroidamplification in primary cells from patients withpolycythemia vera. Blood. 2007;110(3):1013-1021.

23. Pronier E, Almire C, Mokrani H, et al. Inhibition ofTET2-mediated conversion of 5-methylcytosine to5-hydroxymethylcytosine disturbs erythroid andgranulomonocytic differentiation of humanhematopoietic progenitors. Blood. 2011;118(9):2551-2555.

24. Kotecha N, Flores NJ, Irish JM, et al. Single-cellprofiling identifies aberrant STAT5 activation inmyeloid malignancies with specific clinical andbiologic correlates. Cancer Cell. 2008;14(4):335-343.

25. Gaipa G, Bugarin C, Longoni D, et al. AberrantGM-CSF signal transduction pathway in juvenilemyelomonocytic leukemia assayed by flowcytometric intracellular STAT5 phosphorylationmeasurement. Leukemia. 2009;23(4):791-793.

26. Ramshaw HS, Bardy PG, Lee MA, et al. Chronicmyelomonocytic leukemia requires granulocyte-macrophage colony-stimulating factor for growthin vitro and in vivo. Exp Hematol. 2002;30(10):1124-1131.

27. Smith AE, Mohamedali AM, Kulasekararaj A,et al. Next-generation sequencing of the TET2gene in 355 MDS and CMML patients reveals low-abundance mutant clones with early origins, but

indicates no definite prognostic value. Blood.2010;116(19):3923-3932.

28. Langemeijer SM, Kuiper RP, Berends M, et al.Acquired mutations in TET2 are common inmyelodysplastic syndromes. Nat Genet. 2009;41(7):838-842.

29. Loh ML. Recent advances in the pathogenesisand treatment of juvenile myelomonocyticleukaemia. Br J Haematol. 2011;152(6):677-687.

30. Nowell PC. The clonal evolution of tumor cellpopulations. Science. 1976;194(4260):23-28.

31. Notta F, Mullighan CG, Wang JC, et al.Evolution of human BCR-ABL1 lymphoblasticleukaemia-initiating cells. Nature. 2011;469(7330):362-367.

32. Damm F, Fontenay M, Bernard OA. Pointmutations in myelodysplastic syndromes. N Engl JMed. 2011;365(12):1154-1155, author reply 155.

33. Ashworth A, Lord CJ, Reis-Filho JS. Geneticinteractions in cancer progression and treatment.Cell. 2011;145(1):30-38.

34. Stein BL, Williams DM, Rogers O, et al. Diseaseburden at the progenitor level is a feature of primarymyelofibrosis: a multivariable analysis of 164 JAK2V617F-positive myeloproliferative neoplasmpatients. Exp Hematol. 2011;39(1):95-101.

35. Jan M, Snyder TM, Corces-Zimmerman MR, et al.Clonal evolution of preleukemic hematopoieticstem cells precedes human acute myeloidleukemia. Sci Transl Med. 2012;4(149):149ra118.

36. Busque L, Patel JP, Figueroa ME, et al. Recurrentsomatic TET2 mutations in normal elderlyindividuals with clonal hematopoiesis. Nat Genet.2012;44(11):1179-1181.

37. Navin N, Kendall J, Troge J, et al. Tumour evolutioninferred by single-cell sequencing. Nature. 2011;472(7341):90-94.

38. Hou Y, Song L, Zhu P, et al. Single-cell exomesequencing and monoclonal evolution of a JAK2-negative myeloproliferative neoplasm. Cell. 2012;148(5):873-885.

39. Sarrazin S, Mossadegh-Keller N, Fukao T, et al.MafB restricts M-CSF-dependent myeloidcommitment divisions of hematopoietic stem cells.Cell. 2009;138(2):300-313.

40. Rieger MA, Hoppe PS, Smejkal BM, et al.Hematopoietic cytokines can instruct lineage choice.Science. 2009;325(5937):217-218.

41. Ko M, Bandukwala HS, An J, et al. Ten-Eleven-Translocation 2 (TET2) negatively regulateshomeostasis and differentiation of hematopoieticstem cells in mice. Proc Natl Acad Sci USA. 2011;108(35):14566-14571.

42. Quivoron C, Couronne L, Della Valle V, et al.TET2 inactivation results in pleiotropichematopoietic abnormalities in mouse and is arecurrent event during human lymphomagenesis.Cancer Cell. 2011;20(1):25-38.

43. Elliott MA, Verstovsek S, Dingli D, et al.Monocytosis is an adverse prognostic factorfor survival in younger patients with primarymyelofibrosis. Leuk Res. 2007;31(11):1503-1509.

BLOOD, 21 MARCH 2013 x VOLUME 121, NUMBER 12 CLONAL ARCHITECTURE IN CMML 2197

For personal use only.on May 30, 2017. by guest www.bloodjournal.orgFrom

44. Muller-Sieburg CE, Sieburg HB, Bernitz JM, et al.Stem cell heterogeneity: implications for agingand regenerative medicine. Blood. 2012;119(17):3900-3907.

45. Wang SA, Galili N, Cerny J, et al. Chronicmyelomonocytic leukemia evolving from

preexisting myelodysplasia shares many featureswith de novo disease. Am J Clin Pathol. 2006;126(5):789-797.

46. Boiocchi L, Espinal-Witter R, Geyer JT, et al.Development of monocytosis in patientswith primary myelofibrosis indicates an

accelerated phase of the disease. Mod Pathol.2013;26(2):204-212.

47. Goardon N, Marchi E, Atzberger A, et al.Coexistence of LMPP-like and GMP-likeleukemia stem cells in acute myeloidleukemia. Cancer Cell. 2011;19(1):138-152.

2198 ITZYKSON et al BLOOD, 21 MARCH 2013 x VOLUME 121, NUMBER 12

For personal use only.on May 30, 2017. by guest www.bloodjournal.orgFrom

online January 14, 2013 originally publisheddoi:10.1182/blood-2012-06-440347

2013 121: 2186-2198  

Vainchenker, Nathalie Droin and Eric SolaryMeurice, Cédric Oréar, François Delhommeau, Olivier A. Bernard, Michaela Fontenay, WilliamBerthon, Lionel Adès, Pierre Fenaux, Uwe Platzbecker, Olivier Gagey, Philippe Rameau, Guillaume Raphaël Itzykson, Olivier Kosmider, Aline Renneville, Margot Morabito, Claude Preudhomme, Céline Clonal architecture of chronic myelomonocytic leukemias 

http://www.bloodjournal.org/content/121/12/2186.full.htmlUpdated information and services can be found at:

(494 articles)Plenary Papers    (1674 articles)Myeloid Neoplasia   

Articles on similar topics can be found in the following Blood collections

http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requestsInformation about reproducing this article in parts or in its entirety may be found online at:

http://www.bloodjournal.org/site/misc/rights.xhtml#reprintsInformation about ordering reprints may be found online at:

http://www.bloodjournal.org/site/subscriptions/index.xhtmlInformation about subscriptions and ASH membership may be found online at:

  Copyright 2011 by The American Society of Hematology; all rights reserved.of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society

For personal use only.on May 30, 2017. by guest www.bloodjournal.orgFrom