Advertisement

Mechanisms underlying the heterogeneity of myelodysplastic syndromes

  • Charles Dussiau
    Affiliations
    Institut Cochin, Institut National de la Santé et de la Recherche Médicale U1016, Centre National de la Recherche Scientifique, Unité Mixte de Recherche 8104, Université Paris Descartes, and Assistance Publique-Hôpitaux de Paris, Hôpitaux Universitaires Paris Centre, Service d'Hématologie Biologique, Paris, France
    Search for articles by this author
  • Michaela Fontenay
    Correspondence
    Offprint requests to:Hôpitaux Universitaires Paris Centre, Laboratoire d'Hématologie, Bâtiment Jean Dausset, Hôpital Cochin, 27, rue du Faubourg Saint-Jacques, 75014, Paris, France;
    Affiliations
    Institut Cochin, Institut National de la Santé et de la Recherche Médicale U1016, Centre National de la Recherche Scientifique, Unité Mixte de Recherche 8104, Université Paris Descartes, and Assistance Publique-Hôpitaux de Paris, Hôpitaux Universitaires Paris Centre, Service d'Hématologie Biologique, Paris, France
    Search for articles by this author
Open AccessPublished:November 22, 2017DOI:https://doi.org/10.1016/j.exphem.2017.10.004

      Highlights

      • In myelodysplastic syndromes (MDSs), the disease-initiating cell is a rare hematopoietic stem cell (HSC) that transmits the genetic abnormalities to its myeloid and lymphoid progeny.
      • The heterogeneity of MDS phenotypes is related to the diversity of genetics, the various combinations and order of mutations in cooperating genes, and the variegation of clonal hematopoietic hierarchy.
      • A mutation in a granulomonocytic progenitor may mark the transformation into acute myeloid leukemia.
      Myelodysplastic syndromes (MDS) are hematopoietic stem cell (HSC) disorders in which recurrent chromosome abnormalities and gene mutations define a clonal hematopoiesis. The MDS-initiating cell is a rare HSC which transmits the genetic abnormalities to its myeloid and lymphoid progeny. The heterogeneity of MDS phenotypes could be linked to the diversity of genetic events involving epigenetic regulators, chromatin modifiers, splicing factors, transcription factors and signaling adaptors, the various combinations and order of mutations in cooperating genes, and the variegation of clonal hematopoietic hierarchy. Usually, epigenetic and splicing gene mutations occur first. A combination of one epigenetic event with a splicing gene alteration is frequent. The HSC compartment is invaded by a dominant and few minor clones organized linearly or with a branched architecture. The dominant clone containing the first initiating mutations produces myeloid and lymphoid lineages in transplanted immune-deficient mice. The mutations confer a selective advantage to myeloid progenitors at the expense of lymphoid progenitors. In the context of differentiation, one mutation may favor the amplification of granulo-monocytic progenitor, which drives the transformation into acute myeloid leukemia. Understanding the hierarchy of mutations provides insights on the mechanism of transformation. Investigation of mutation pattern and distribution along the hematopoietic tree may influence the therapeutic decision for targeted therapy.
      Myelodysplastic syndromes (MDSs) are hematopoietic stem cell (HSC) disorders in which recurrent chromosome abnormalities and gene mutations define a clonal hematopoiesis. The MDS-initiating cell is a rare HSC that transmits the genetic abnormalities to its myeloid and lymphoid progeny. The heterogeneity of MDS phenotypes could be linked to the diversity of genetic events involving epigenetic regulators, chromatin modifiers, splicing factors, transcription factors and signaling adaptors, the various combinations and order of mutations in cooperating genes, and the variegation of clonal hematopoietic hierarchy. Usually, epigenetic and splicing gene mutations occur first. A combination of one epigenetic event with a splicing gene alteration is frequent. The HSC compartment is invaded by a dominant and few minor clones organized linearly or with a branched architecture. The dominant clone containing the first initiating mutations produces myeloid and lymphoid lineages in transplanted immune-deficient mice. The mutations confer a selective advantage to myeloid progenitors at the expense of lymphoid progenitors. In the context of differentiation, one mutation may favor the amplification of granulomonocytic progenitor, which drives the transformation into acute myeloid leukemia. Understanding the hierarchy of mutations provides insights on the mechanism of transformation. Investigation of mutation pattern and distribution along the hematopoietic tree may influence the therapeutic decision for targeted therapy.
      Myelodysplastic syndromes (MDS) are common myeloid malignancies of the elderly. These syndromes are characterized by ineffective hematopoiesis and a propensity to develop an acute myeloid leukemia (AML) [
      • Corey S.J.
      • Minden M.D.
      • Barber D.L.
      • Kantarjian H.
      • Wang J.C.
      • Schimmer A.D.
      Myelodysplastic syndromes: the complexity of stem-cell diseases.
      ]. The diversity of disease subtypes is described in the 2016 World Health Organization (WHO) classification [
      • Arber D.A.
      • Orazi A.
      • Hasserjian R.
      • et al.
      The 2016 revision to the World Health Organization classification of myeloid neoplasms and acute leukemia.
      ]. Acquisition of genomic abnormalities drives the transformation of a normal hematopoietic cell into a clonal cell. The type of mutation and the combination of genetic events may be responsible for the heterogeneity of phenotypes and clinical outcomes. Clonal hematopoiesis is defined by the presence of genetic events such as chromosome rearrangements and gene alterations. Losses, gains, and translocations that usually affect chromosomes 3, 5, 7, 8, 11, 12, 17, 19, 20, X, and Y are not always specific for these diseases. This led the WHO committee to retain abnormalities of chromosomes 5, 7, 11, 12, 13, 17, and X as MDS-related abnormalities [
      • Arber D.A.
      • Orazi A.
      • Hasserjian R.
      • et al.
      The 2016 revision to the World Health Organization classification of myeloid neoplasms and acute leukemia.
      ] (Table 1). Alterations such as nucleotide changes, deletions or insertions in genes involved in epigenetic modifications of DNA or histones, splicing, transcription, and signaling [
      • Delhommeau F.
      • Dupont S.
      • Della Valle V.
      • et al.
      Mutation in TET2 in myeloid cancers.
      ,
      • Gelsi-Boyer V.
      • Trouplin V.
      • Adélaïde J.
      • et al.
      Mutations of polycomb-associated gene ASXL1 in myelodysplastic syndromes and chronic myelomonocytic leukaemia.
      ,
      • Ernst T.
      • Chase A.J.
      • Score J.
      • et al.
      Inactivating mutations of the histone methyltransferase gene EZH2 in myeloid disorders.
      ,
      • Yoshida K.
      • Sanada M.
      • Shiraishi Y.
      • et al.
      Frequent pathway mutations of splicing machinery in myelodysplasia.
      ] have been reported extensively (Table 2). These mutations are not restricted to myeloid diseases but are also seen in lymphoid malignancies [
      • Quivoron C.
      • Couronné L.
      • Della Valle V.
      • et al.
      TET2 inactivation results in pleiotropic hematopoietic abnormalities in mouse and is a recurrent event during human lymphomagenesis.
      ,
      • Wang L.
      • Lawrence M.S.
      • Wan Y.
      • et al.
      SF3B1 and other novel cancer genes in chronic lymphocytic leukemia.
      ]. The mechanisms of tumor heterogeneity will be depicted with a special attention to the diversity of genetic events, the variegation of mutational landscape in the hematopoietic hierarchy, and the clonal evolution during natural course of the disease or under therapy.
      Table 1Cytogenetic abnormalities related to myelodysplasia according to the WHO classification of 2016
      TypesChromosomal abnormalities
      Complex Karyotype (three or more abnormalities)
      Unbalanced Abnormalities–7/del(7q)

      del(5q)/t(5q)

      i(17q)/t(17p)

      –13/del(13q)

      del(11q)

      del(12p)/t(12p)

      Idic(X)(q13)
      Balanced Abnormalitiest(11;16)(q23.3;p13.3)

      t(3;21)(q26.2;q22.1)

      t(1;3)(p36.3;q21.2)

      t(2;11)(p21;q23.3)

      t(5;12)(q32;p13.2)

      t(5;7)(q32;q11.2)

      t(5;17)(q32;p13.2)

      t(5;10)(q32;q21.2)

      t(3;5)(q25.3;q35.1)
      Table 2Recurrent mutations in MDS
      GenesSymbols
      Epigenetic RegulatorsTET2, DNMT3A, IDH2, IDH1
      Splicing FactorsSF3B1, SRSF2, U2AF1, ZRSR2, PHF6
      Chromatin ModifiersASXL1, EZH2, MLL2, KDM6A, ATRX
      Transcription Factors and CofactorsRUNX1, TP53, BCOR, GATA2, CUX1, EP300, CREBBP, WT1, IRF1, ETV6
      SignalingCBL, NRAS, JAK2, KRAS, KIT, MPL, PTPN11, NF1, CDKN2A
      CohesinsSTAG2, RAD21
      OtherNPM1

      Onset of clonal hematopoiesis

      The Darwinian model of evolution is convoked to explain cancer development, implicating the onset of random mutations in a cancer-initiating cell and their selection. Primitive tumors demonstrate a variable prevalence of genetic events with the lowest prevalence in hematological malignancies compared with solid tumors such as melanoma or lung carcinoma, which are mostly secondary to toxic exposure [
      • Alexandrov L.B.
      • Nik-Zainal S.
      • Wedge D.C.
      • et al.
      Signatures of mutational processes in human cancer.
      ]. The bone marrow could be assimilated to a sanctuary relatively protected from environmental toxics. In addition, the hierarchical organization of hematopoiesis may protect the cells from a clonal evolution [
      • Pepper J.W.
      • Sprouffske K.
      • Maley C.C.
      Animal cell differentiation patterns suppress somatic evolution.
      ]. Mutations accumulate during decades, so the number of mutations in self-renewing tissues could be correlated directly with age [
      • Tomasetti C.
      • Vogelstein B.
      • Parmigiani G.
      Half or more of the somatic mutations in cancers of self-renewing tissues originate prior to tumor initiation.
      ]. A mutational signature of aging defined by the accumulation of C > T or T > C transitions is highly predominant in MDS and is a plausible consequence of DNA damage that accumulates in cycling hematopoietic stem cells (HSCs) [
      • Alexandrov L.B.
      • Nik-Zainal S.
      • Wedge D.C.
      • et al.
      Signatures of mutational processes in human cancer.
      ,
      • Walter D.
      • Lier A.
      • Geiselhart A.
      • et al.
      Exit from dormancy provokes DNA-damage-induced attrition in haematopoietic stem cells.
      ].
      The presence of mutations in leukemia and/or lymphoma-associated genes can arise without evidence of hematological disorder. A pioneer study from Busque et al. reported mutations in the TET2 gene in women older than 60 years [
      • Busque L.
      • Patel J.P.
      • Figueroa M.E.
      • et al.
      Recurrent somatic TET2 mutations in normal elderly individuals with clonal hematopoiesis.
      ]. Whole-exome sequencing of peripheral blood samples in large cohorts of individuals confirmed that mutation frequency increases with age and affects more than 20% of individuals older than 90 years [
      • Xie M.
      • Lu C.
      • Wang J.
      • et al.
      Age-related mutations associated with clonal hematopoietic expansion and malignancies.
      ,
      • Jaiswal S.
      • Fontanillas P.
      • Flannick J.
      • et al.
      Age-related clonal hematopoiesis associated with adverse outcomes.
      ,
      • Genovese G.
      • Kähler A.K.
      • Handsaker R.E.
      • et al.
      Clonal hematopoiesis and blood-cancer risk inferred from blood DNA sequence.
      ,
      • McKerrell T.
      • Park N.
      • Moreno T.
      • et al.
      Leukemia-associated somatic mutations drive distinct patterns of age-related clonal hematopoiesis.
      ]. The most common recurrently mutated genes are DNMT3A, TET2, ASXL1, TP53, and JAK2. A new entity called clonal hematopoiesis of indeterminate potential (CHIP) has been defined based on the presence of mutations (at least 2% variant allele frequency in peripheral blood) in the absence of dysplasia or cytopenia [
      • Steensma D.P.
      • Bejar R.
      • Jaiswal S.
      • et al.
      Clonal hematopoiesis of indeterminate potential and its distinction from myelodysplastic syndromes.
      ]. A low level of mutations can also be found in a fraction of patients with moderate dysplasia or idiopathic cytopenias of undetermined significance (ICUS), and this may represent a transition to MDS, a positive diagnosis of which is based on morphological WHO criteria, clinical evolution, and variant allele frequency over 10% [
      • Steensma D.P.
      • Bejar R.
      • Jaiswal S.
      • et al.
      Clonal hematopoiesis of indeterminate potential and its distinction from myelodysplastic syndromes.
      ]. Preclinical clonal hematopoiesis seems to indicate a high risk of hematological malignancy development and disease progression. The subjects have a decreased overall survival compared with age-matched controls [
      • Jaiswal S.
      • Fontanillas P.
      • Flannick J.
      • et al.
      Age-related clonal hematopoiesis associated with adverse outcomes.
      ,
      • Genovese G.
      • Kähler A.K.
      • Handsaker R.E.
      • et al.
      Clonal hematopoiesis and blood-cancer risk inferred from blood DNA sequence.
      ,
      • Cargo C.A.
      • Rowbotham N.
      • Evans P.A.
      • et al.
      Targeted sequencing identifies patients with preclinical MDS at high risk of disease progression.
      ]. Prospective studies of patients devoid of disease-modifying treatments are required to evaluate the evolution of CHIP. Related to the evolution from CHIP to MDS and secondary AML after MDS is the occurrence of mutations in cooperating genes leading to functional defects of affected cells and their deleterious interaction with the bone marrow microenvironment.
      Unexpectedly regarding the incomplete penetrance of dysplastic features, whole-genome sequencing of MDS bone marrow samples has revealed the presence of a dominant clone in which mutations are expressed in more than 90% of the cells coexisting with few minor clones. Those mutations segregate in a mean of two clones in early MDS and in more than three clones at the time of AML transformation [
      • Walter M.J.
      • Shen D.
      • Ding L.
      • et al.
      Clonal architecture of secondary acute myeloid leukemia.
      ]. Ten to 20 mutations are detected in individual samples by whole-exome sequencing, confirming the oligoclonal nature of the bone marrow [
      • Yoshida K.
      • Sanada M.
      • Shiraishi Y.
      • et al.
      Frequent pathway mutations of splicing machinery in myelodysplasia.
      ,
      • Mossner M.
      • Jann J.C.
      • Wittig J.
      • et al.
      Mutational hierarchies in myelodysplastic syndromes dynamically adapt and evolve upon therapy response and failure.
      ,
      • da Silva-Coelho P.
      • Kroeze L.I.
      • Yoshida K.
      • et al.
      Clonal evolution in myelodysplastic syndromes.
      ]. A median of four to six mutations affect genes that are mutated recurrently in hematological malignancies; that is, epigenetic regulators, splicing factors, chromatin modifiers, transcription factors, cohesins, and signaling adaptors. Analysis of sequential samples allows backtracking the initiating events [
      • Mossner M.
      • Jann J.C.
      • Wittig J.
      • et al.
      Mutational hierarchies in myelodysplastic syndromes dynamically adapt and evolve upon therapy response and failure.
      ,
      • da Silva-Coelho P.
      • Kroeze L.I.
      • Yoshida K.
      • et al.
      Clonal evolution in myelodysplastic syndromes.
      ]. In most cases, the founder genetic lesions target recurrently epigenetic genes involved in the regulation of DNA methylation, such as TET2 or DNMT3A; genes involved in the remodeling of chromatin, such as ASXL1 or EZH2; and genes involved in RNA splicing, such as SF3B1, SRSF2, U2AF1, or ZRSR2 [
      • Delhommeau F.
      • Dupont S.
      • Della Valle V.
      • et al.
      Mutation in TET2 in myeloid cancers.
      ,
      • Gelsi-Boyer V.
      • Trouplin V.
      • Adélaïde J.
      • et al.
      Mutations of polycomb-associated gene ASXL1 in myelodysplastic syndromes and chronic myelomonocytic leukaemia.
      ,
      • Ernst T.
      • Chase A.J.
      • Score J.
      • et al.
      Inactivating mutations of the histone methyltransferase gene EZH2 in myeloid disorders.
      ,
      • Yoshida K.
      • Sanada M.
      • Shiraishi Y.
      • et al.
      Frequent pathway mutations of splicing machinery in myelodysplasia.
      ]. Mutations affecting cell differentiation or proliferation pathways (RUNX1, GATA2, BCOR, N/KRAS, CBL) or cohesins (STAG2, RAD21) more likely occur during the progression from MDS to AML [
      • Damm F.
      • Chesnais V.
      • Nagata Y.
      • et al.
      BCOR and BCORL1 mutations in myelodysplastic syndromes and related disorders.
      ,
      • Mazumdar C.
      • Shen Y.
      • Xavy S.
      • et al.
      Leukemia-associated cohesin mutants dominantly enforce stem cell programs and impair human hematopoietic progenitor differentiation.
      ]. These secondary mutations usually accumulate in cells that exhibit a genetic background constituted by mutations present at the onset of the disease [
      • Damm F.
      • Chesnais V.
      • Nagata Y.
      • et al.
      BCOR and BCORL1 mutations in myelodysplastic syndromes and related disorders.
      ,
      • Mazumdar C.
      • Shen Y.
      • Xavy S.
      • et al.
      Leukemia-associated cohesin mutants dominantly enforce stem cell programs and impair human hematopoietic progenitor differentiation.
      ,
      • Malcovati L.
      • Karimi M.
      • Papaemmanuil E.
      • et al.
      SF3B1 mutation identifies a distinct subset of myelodysplastic syndrome with ring sideroblasts.
      ]. Secondary AML post-MDS and de novo AML share mutations in DNMT3A, TET2, IDH1/2, and TP53 genes. In contrast, mutations in NPM1, FLT3, and MLL/KTM2A and chromosomal rearrangements of the core binding factor are usually absent in MDS.
      The order of mutation could contribute to the phenotypic presentation of the disease. This has been emphasized in myeloproliferative neoplasms, in which the initiating event may be either a TET2 or a JAK2 mutation. Patients presenting with a TET2 mutation as a first event are older than patients with a JAK2 mutation as a first event. The bone marrow hierarchy is different, with more common myeloid progenitors (CMPs) in “TET2-first” and more erythroid and megakaryocytic progenitors in “JAK2-first” cases. Patients with a JAK2 mutation at the beginning present with polycythemia vera more likely than essential thrombocythemia and have an increased risk of thrombosis [
      • Ortmann C.A.
      • Kent D.G.
      • Nangalia J.
      • et al.
      Effect of mutation order on myeloproliferative neoplasms.
      ]. In MDS, the influence of mutation order on the patient phenotype is less clear. However, even if the two mutations are clonal, meaning that they affect more or less 50% of bone marrow cells, the clone that drives the disease in immunodeficient mice is the clone that contains the earliest events. In two-thirds of patients with an SF3B1 mutation, this event occurs before mutations in DNMT3A or TET2. In this case, the dominant SF3B1-mutated clone of the hematopoietic stem/progenitor cell (HSPC) compartment gives an SF3B1-mutated progeny of myeloid and lymphoid cells in xenograft experiments. In contrast, when the mutation SF3B1 occurs after the epigenetic event, and even if SF3B1 mutation belongs to a dominant clone, it appears as a minor myeloid clone in mouse bone marrow [
      • Chesnais V.
      • Arcangeli M.L.
      • Delette C.
      • et al.
      Architectural and functional heterogeneity of hematopoietic stem/progenitor cells in non-del(5q) myelodysplastic syndromes.
      ]. The companion mutations in the different clones may influence engraftment capacities and cooperate with the initiating events to propagate the disease. The type of founder mutation may also influence the phenotype. The diagnosis of sideroblastic anemia is considered as soon as the SF3B1 mutation is detected, even if the percentage of bone marrow ring sideroblasts is below 15%. There is a significant correlation between SF3B1 variant allele burden and the percentage of ring sideroblasts [
      • Malcovati L.
      • Karimi M.
      • Papaemmanuil E.
      • et al.
      SF3B1 mutation identifies a distinct subset of myelodysplastic syndrome with ring sideroblasts.
      ]. Indeed, the knock-in of SF3B1K700E in mice induces a macrocytic anemia and an increased expansion of long-term HSCs (LT-HSCs) [
      • Obeng E.A.
      • Chappell R.J.
      • Seiler M.
      • et al.
      Physiologic expression of sf3b1(k700e) causes impaired erythropoiesis, aberrant splicing, and sensitivity to therapeutic spliceosome modulation.
      ]. Furthermore, SF3B1-mutated HSCs isolated from patients with a sideroblastic anemia and injected to immunodeficient mice undergo abnormal erythroid differentiation assessed by the onset of ring sideroblasts at the polychromatophilic stage [
      • Mortera-Blanco T.
      • Dimitriou M.
      • Woll P.S.
      • et al.
      SF3B1-initiating mutations in MDS-RSs target lymphomyeloid hematopoietic stem cells.
      ]. SF3B1 mutations are never detected in CHIP. This suggests that the onset of SF3B1 mutation drives the entry into a symptomatic disease. SF3B1 mutation is rarely isolated, but rather is usually associated with a DNMT3A or TET2 mutation. However, the co-occurrence of an epigenetic event does not prevent the formation of ring sideroblasts. This also suggests that SF3B1 and TET2 or DNMT3A do not act as epistatic genes. Targeted resequencing using next-generation sequencing often fails to ascertain the order of mutations in the dominant clone. Single-cell genotyping is required to identify the first initiating event among the diversity of combinations. Therefore, the question of the impact of mutation order on the phenotype of MDS remains an open question.

      Normal hematopoietic hierarchy

      The recent contributions to the understanding of normal hematopoiesis help in deciphering the mechanisms of initiation and propagation of myeloid malignancies, which mainly result from the insult of the HSCs. At the top of hematopoietic hierarchy, the HSC demonstrates self-renewal activity and differentiation capacities toward both myeloid and lymphoid lineages. The capability of HSCs to reconstitute human hematopoiesis beyond 12 weeks with mature cells of different types into irradiated immunodeficient mice defines the LT-HSC. Its immediate progeny, the short-term HSCs (ST-HSCs), support a transient multilineage engraftment in mice. Phenotypically, CD34+CD38–/loCD45RA LT-HSCs or ST-HSCs are distinguished by the expression of CD90/Thy1, which is lost by ST-HSCs [
      • Majeti R.
      • Park C.Y.
      • Weissman I.L.
      Identification of a hierarchy of multipotent hematopoietic progenitors in human cord blood.
      ,
      • Mazurier F.
      • Doedens M.
      • Gan O.I.
      • Dick J.E.
      Rapid myeloerythroid repopulation after intrafemoral transplantation of NOD-SCID mice reveals a new class of human stem cells.
      ]. Membrane expression of integrin α6, CD49f, also discriminates CD49f+ LT-HSCs from CD49f ST-HSCs. The self-renewing compartment of CD34+CD38CD45RACD90/Thy1 cells extends to the multipotent progenitors (MPPs).
      Single HSC transplantation experiments have demonstrated that the maintenance of the graft and the reconstitution of myeloid and lymphoid lineages is variable, suggesting that the HSC compartment is highly heterogeneous [
      • Dykstra B.
      • Kent D.
      • Bowie M.
      • et al.
      Long-term propagation of distinct hematopoietic differentiation programs in vivo.
      ,
      • Yamamoto R.
      • Morita Y.
      • Ooehara J.
      • et al.
      Clonal analysis unveils self-renewing lineage-restricted progenitors generated directly from hematopoietic stem cells.
      ]. In particular, these experiments identified a myeloid-restricted progenitor with long-term repopulating activity that produces either a megakaryocytic and megakaryocytic-erythroid progeny or a common myeloid progeny [
      • Yamamoto R.
      • Morita Y.
      • Ooehara J.
      • et al.
      Clonal analysis unveils self-renewing lineage-restricted progenitors generated directly from hematopoietic stem cells.
      ]. The heterogeneity of the HSC compartment might be explained by the sequential loss of lineage potential by differentiating HSCs or the coexistence of distinct subsets of HSCs with multilineage or unilineage potential. MPPs with various lineage potentials have been identified downstream of the HSC and are separated in several classes (MPP1, MPP2, MMP3, and MPP4) on the basis of phenotypic, epigenomic, transcriptomic, proteomic, and functional characteristics [
      • Cabezas-Wallscheid N.
      • Klimmeck D.
      • Hansson J.
      • et al.
      Identification of regulatory networks in HSCs and their immediate progeny via integrated proteome, transcriptome, and DNA methylome analysis.
      ,
      • Paul F.
      • Arkin Y.
      • Giladi A.
      • et al.
      Transcriptional heterogeneity and lineage commitment in myeloid progenitors.
      ,
      • Perié L.
      • Duffy K.R.
      • Kok L.
      • de Boer R.J.
      • Schumacher T.N.
      The branching point in erythro-myeloid differentiation.
      ,
      • Notta F.
      • Zandi S.
      • Takayama N.
      • et al.
      Distinct routes of lineage development reshape the human blood hierarchy across ontogeny.
      ]. The most abundant MPPs are the lymphoid-primed MPPs (LMPPs or MPP4). This is a lineage-biased population with low myeloid potential and high lymphoid potential controlling the fate of T, B, and natural killer lymphocytes and monocytes [
      • Naik S.H.
      • Perié L.
      • Swart E.
      • et al.
      Diverse and heritable lineage imprinting of early haematopoietic progenitors.
      ,
      • Doulatov S.
      • Notta F.
      • Eppert K.
      • Nguyen L.T.
      • Ohashi P.S.
      • Dick J.E.
      Revised map of the human progenitor hierarchy shows the origin of macrophages and dendritic cells in early lymphoid development.
      ,
      • Laurenti E.
      • Doulatov S.
      • Zandi S.
      • et al.
      The transcriptional architecture of early human hematopoiesis identifies multilevel control of lymphoid commitment.
      ]. However, this cell retains both lymphoid and myeloid transcriptomic programs, whereas its epigenome is oriented toward the lymphoid lineage by inactivating DNA methylation of myeloid enhancer regions and not lymphoid enhancers [
      • Yu V.W.
      • Yusuf R.Z.
      • Oki T.
      • et al.
      Epigenetic memory underlies cell-autonomous heterogeneous behavior of hematopoietic stem cells.
      ]. MPP1 cells have a multilineage myelo-erythro-megakaryocytic (My-Er-Mk) potential and a lymphoid potential in primary transplantation that is lost by secondary transplantation into mice [
      • Cabezas-Wallscheid N.
      • Klimmeck D.
      • Hansson J.
      • et al.
      Identification of regulatory networks in HSCs and their immediate progeny via integrated proteome, transcriptome, and DNA methylome analysis.
      ]. MPP2 and MPP3 have a quite exclusive myeloid orientation with an Er-Mk potential for MPP2 and a granulomonocytic potential for MPP3 [
      • Paul F.
      • Arkin Y.
      • Giladi A.
      • et al.
      Transcriptional heterogeneity and lineage commitment in myeloid progenitors.
      ,
      • Notta F.
      • Zandi S.
      • Takayama N.
      • et al.
      Distinct routes of lineage development reshape the human blood hierarchy across ontogeny.
      ,
      • Pietras E.M.
      • Reynaud D.
      • Kang Y.-A.
      • et al.
      Functionally distinct subsets of lineage-biased multipotent progenitors control blood production in normal and regenerative conditions.
      ]. According to Notta et al., MPPs with multilineage potential CD34+CD38CD45RACD90/Thy1CD71MPL can be distinguished from MPPs with Er-Mk or Er potential on the basis of CD71 expression or CD71 and MPL expression, respectively [
      • Notta F.
      • Zandi S.
      • Takayama N.
      • et al.
      Distinct routes of lineage development reshape the human blood hierarchy across ontogeny.
      ]. The proposed model in mice is that two distinct populations of MPPs—MPP2 and MPP3—are myeloid specified and one LMPP/MPP4 supplies the lymphoid lineage [
      • Pietras E.M.
      • Reynaud D.
      • Kang Y.-A.
      • et al.
      Functionally distinct subsets of lineage-biased multipotent progenitors control blood production in normal and regenerative conditions.
      ].
      The multipotent cell populations are in a poised state from which they exit for lineage commitment driven by exogenous signals. This state could be governed by low levels of transcription factors under a threshold required for lineage commitment. By tracking fluorochrome-tagged GATA-1 and PU-1 transcription factors at the single-cell level, Hoppe et al. demonstrated that no cell, even putatively multipotent cells, coexpress these transcription factors [
      • Hoppe P.S.
      • Schwarzfischer M.
      • Loeffler D.
      • et al.
      Early myeloid lineage choice is not initiated by random PU.1 to GATA1 protein ratios.
      ]. Progenitors express one transcription factor that drives the lineage commitment. In the classical model of hematopoiesis, the first branching point arising at the level of committed CD34+CD38+ progenitors splits lymphoid and myeloid lineages into common lymphoid progenitors (CLPs) and CMPs, which then separate into megakaryocytic–erythroid progenitors (MEPs) and granulomonocytic progenitors (GMPs). Fed by the upstream multilineage or unilineage-specified MPPs, the CD34+CD38+CD45RAFlt3+ CMPs and the CD34+CD38+CD45RAFlt3 MEPs have a predominant unilineage outcome: myeloid and erythroid, respectively [
      • Notta F.
      • Zandi S.
      • Takayama N.
      • et al.
      Distinct routes of lineage development reshape the human blood hierarchy across ontogeny.
      ].
      In basal conditions, blood homeostasis is maintained by long-survivor myeloid progenitors without the contribution of HSCs, which are mostly quiescent [
      • Sun J.
      • Ramos A.
      • Chapman B.
      • et al.
      Clonal dynamics of native haematopoiesis.
      ]. The feeding role is devoted to MPPs. HSCs stay out of the cell cycle, which means that they are not prone to acquiring mutations in case of DNA misrepair during mitosis. In regenerating conditions, HSCs self-renew to feed the MPP compartment transiently. LMPP/MPP4 cells can be reprogrammed toward the myeloid lineage to participate in the bone marrow reconstitution [
      • Pietras E.M.
      • Reynaud D.
      • Kang Y.-A.
      • et al.
      Functionally distinct subsets of lineage-biased multipotent progenitors control blood production in normal and regenerative conditions.
      ]. HSCs differentiate into MPP2 and MPP3 and then into GMPs that organize in clusters upon stimulation by cytokines stem cell factor (SCF) and granulocyte colony-stimulating factor (G-CSF). This process is arrested by the release of inhibitory factors by neighboring megakaryocytes, which restore HSC quiescence [
      • Hérault A.
      • Binnewies M.
      • Leong S.
      • et al.
      Myeloid progenitor cluster formation drives emergency and leukaemic myelopoiesis.
      ].

      HSC is targeted by genetic events

      Infection, inflammation, or chronic blood loss are physiological stresses that may instruct HSCs to exit from quiescence. When cycling for self-renewal or asymmetrical division, HSCs and progenitors are more sensitive to genetic alterations than more differentiated cells. In these conditions, HSCs are submitted to collapse of mitochondrial membrane potential, production of reactive oxygen species, and DNA damage. Repeated stresses may provoke their attrition [
      • Walter D.
      • Lier A.
      • Geiselhart A.
      • et al.
      Exit from dormancy provokes DNA-damage-induced attrition in haematopoietic stem cells.
      ,
      • Yahata T.
      • Takanashi T.
      • Muguruma Y.
      • et al.
      Accumulation of oxidative DNA damage restricts the self-renewal capacity of human hematopoietic stem cells.
      ]. The functional potential of HSCs also changes during aging. Old HSCs exhibit a reduced capacity to generate their progeny and an altered lymphoid differentiation. In addition, lymphoid-biased HSCs should have a longer lifetime and lower capacities to proliferate than myeloid-biased HSCs [
      • Doulatov S.
      • Notta F.
      • Eppert K.
      • Nguyen L.T.
      • Ohashi P.S.
      • Dick J.E.
      Revised map of the human progenitor hierarchy shows the origin of macrophages and dendritic cells in early lymphoid development.
      ,
      • Laurenti E.
      • Doulatov S.
      • Zandi S.
      • et al.
      The transcriptional architecture of early human hematopoiesis identifies multilevel control of lymphoid commitment.
      ,
      • Pietras E.M.
      • Reynaud D.
      • Kang Y.-A.
      • et al.
      Functionally distinct subsets of lineage-biased multipotent progenitors control blood production in normal and regenerative conditions.
      ]. The positive selection of myeloid-biased HSCs and the alterations in lineage specification programs is beneficial to myeloid differentiation [
      • Beerman I.
      • Bhattacharya D.
      • Zandi S.
      • et al.
      Functionally distinct hematopoietic stem cells modulate hematopoietic lineage potential during aging by a mechanism of clonal expansion.
      ,
      • Pang W.W.
      • Price E.A.
      • Sahoo D.
      • et al.
      Human bone marrow hematopoietic stem cells are increased in frequency and myeloid-biased with age.
      ]. In old mice, HSCs are less functional compared with young mice in the conditions of bone marrow transplantation. Importantly, old HSCs accumulate DNA damage due to limited DNA repair capacities compared with proliferating progenitor cells [
      • Rossi D.J.
      • Bryder D.
      • Seita J.
      • Nussenzweig A.
      • Hoeijmakers J.
      • Weissman I.L.
      Deficiencies in DNA damage repair limit the function of haematopoietic stem cells with age.
      ,
      • Beerman I.
      • Seita J.
      • Inlay M.A.
      • Weissman I.L.
      • Rossi D.J.
      Quiescent hematopoietic stem cells accumulate DNA damage during aging that is repaired upon entry into cell cycle.
      ,
      • Flach J.
      • Bakker S.T.
      • Mohrin M.
      • et al.
      Replication stress is a potent driver of functional decline in ageing haematopoietic stem cells.
      ]. They experiment a replicative stress leading to double-strand DNA breaks in the progeny [
      • Flach J.
      • Bakker S.T.
      • Mohrin M.
      • et al.
      Replication stress is a potent driver of functional decline in ageing haematopoietic stem cells.
      ]. This genotoxic stress may contribute to the onset of mutations and could originate from a deleterious interaction with an inflammatory microenvironment.
      In a mouse model of Shwachman–Diamond syndrome, which mimics a preleukemic syndrome, CD271+ mesenchymal stromal cells produce inflammatory molecules named damage-associated molecular pattern (DAMP) S100A8/9. Recombinant S100A8 induces mitochondrial dysfunction, oxidative stress, and a DNA damage response in HSPCs [
      • Zambetti N.A.
      • Ping Z.
      • Chen S.
      • et al.
      Mesenchymal inflammation drives genotoxic stress in hematopoietic stem cells and predicts disease evolution in human pre-leukemia.
      ]. Interestingly, the expression of S100A8/9 markers is increased in several MDS subtypes, including the 5q– syndrome and therefore could contribute to the induction of a genotoxic stress at the level of MDS HSPCs [
      • Zambetti N.A.
      • Ping Z.
      • Chen S.
      • et al.
      Mesenchymal inflammation drives genotoxic stress in hematopoietic stem cells and predicts disease evolution in human pre-leukemia.
      ,
      • Schneider R.K.
      • Schenone M.
      • Ferreira M.V.
      • et al.
      Rps14 haploinsufficiency causes a block in erythroid differentiation mediated by S100A8 and S100A9.
      ].
      In myeloid malignancies, the clones are thought to emerge from a unique cell that acquired novel mutations sequentially over time, suggesting that the genetic complexity may increase with evolution. Nilsson et al. reported the first evidence for an MDS-initiating cell by identifying a chromosome 5q deletion by fluorescence in situ hybridization in sorted CD34+CD38 cells [
      • Nilsson L.
      • Astrand-Grundström I.
      • Arvidsson I.
      • et al.
      Isolation and characterization of hematopoietic progenitor/stem cells in 5q-deleted myelodysplastic syndromes: evidence for involvement at the hematopoietic stem cell level.
      ]. The cell that is initially targeted by genetic events (either chromosomal deletion or mutation) is a rare HSC [
      • Tehranchi R.
      • Woll P.S.
      • Anderson K.
      • et al.
      Persistent malignant stem cells in del(5q) myelodysplasia in remission.
      ,
      • Woll P.S.
      • Kjällquist U.
      • Chowdhury O.
      • et al.
      Myelodysplastic syndromes are propagated by rare and distinct human cancer stem cells in vivo.
      ]. The del(5q) linlowCD34+CD38CD90/Thy1+CD45RA cells exhibit long-term culture-initiating cell (LTC-IC) activity and capacities to engraft and to expand a myeloid and a lymphoid progeny in immunodeficient NSG mice [
      • Woll P.S.
      • Kjällquist U.
      • Chowdhury O.
      • et al.
      Myelodysplastic syndromes are propagated by rare and distinct human cancer stem cells in vivo.
      ,
      • Thanopoulou E.
      • Cashman J.
      • Kakagianne T.
      • Eaves A.
      • Zoumbos N.
      • Eaves C.
      Engraftment of NOD/SCID-beta2 microglobulin null mice with multilineage neoplastic cells from patients with myelodysplastic syndrome.
      ]. HSC-deriving cells isolated from patients with an SF3B1-mutated sideroblastic anemia are also capable of long-term engraftment [
      • Mortera-Blanco T.
      • Dimitriou M.
      • Woll P.S.
      • et al.
      SF3B1-initiating mutations in MDS-RSs target lymphomyeloid hematopoietic stem cells.
      ,
      • Mian S.A.
      • Rouault-Pierre K.
      • Smith A.E.
      • et al.
      SF3B1 mutant MDS-initiating cells may arise from the haematopoietic stem cell compartment.
      ]. By backtracking the mutations at the single-cell level in the CD34+CD38 HSPC compartment, we have shown that individual CD34+CD38 cells contains 0, 1, 2, 3, or all of the mutations detected in bulk mononuclear cells; are organized in clones that coexist; and are maintained in this compartment. A dominant clone accumulates mutations in a linear succession or a branched architecture downstream of the founding clone or in a separate clone [
      • Chesnais V.
      • Arcangeli M.L.
      • Delette C.
      • et al.
      Architectural and functional heterogeneity of hematopoietic stem/progenitor cells in non-del(5q) myelodysplastic syndromes.
      ,
      • Woll P.S.
      • Kjällquist U.
      • Chowdhury O.
      • et al.
      Myelodysplastic syndromes are propagated by rare and distinct human cancer stem cells in vivo.
      ]. Among HSPCs, LT-HSCs harboring LTC-IC activity contain the mutations of the dominant clone. In xenotransplantation experiments, engrafted cells belonged to the dominant clone and revealed their stem cell nature by yielding both CD15+ or CD14+ myeloid cells and CD19+ lymphoid cells [
      • Mossner M.
      • Jann J.C.
      • Wittig J.
      • et al.
      Mutational hierarchies in myelodysplastic syndromes dynamically adapt and evolve upon therapy response and failure.
      ,
      • Chesnais V.
      • Arcangeli M.L.
      • Delette C.
      • et al.
      Architectural and functional heterogeneity of hematopoietic stem/progenitor cells in non-del(5q) myelodysplastic syndromes.
      ,
      • Woll P.S.
      • Kjällquist U.
      • Chowdhury O.
      • et al.
      Myelodysplastic syndromes are propagated by rare and distinct human cancer stem cells in vivo.
      ].

      Alterations of hematopoietic hierarchy in myeloid malignancies

      In several myeloid malignancies, the hematopoietic tree is highly modified. In aplastic anemia, the CD34+CD38 HSC and MPP compartments are depleted and the CD34+CD38+ compartment is preserved. Among CD34+CD38+ cells, erythroid progenitors are lost and myeloid progenitors are represented normally [
      • Notta F.
      • Zandi S.
      • Takayama N.
      • et al.
      Distinct routes of lineage development reshape the human blood hierarchy across ontogeny.
      ]. It has been demonstrated that the transmission of mutations to the progeny disturbs the steps of development. In BCR-ABL-driven chronic myeloid leukemia (CML), the initiating cell derives from the HSC compartment and drives the relapse [
      • Wang J.C.
      • Lapidot T.
      • Cashman J.D.
      • et al.
      High level engraftment of NOD/SCID mice by primitive normal and leukemic hematopoietic cells from patients with chronic myeloid leukemia in chronic phase.
      ,
      • Corbin A.S.
      • Agarwal A.
      • Loriaux M.
      • Cortes J.
      • Deininger M.W.
      • Druker B.J.
      Human chronic myeloid leukemia stem cells are insensitive to imatinib despite inhibition of BCR-ABL activity.
      ]. During the development of CML, HSCs differentiate into myeloid-biased MPPs. MPP2 and MPP3 feed the GMP compartment rapidly organized in clusters of GMPs that are maintained by the activation of self-renewal pathway and the default of inhibitory cytokines [
      • Hérault A.
      • Binnewies M.
      • Leong S.
      • et al.
      Myeloid progenitor cluster formation drives emergency and leukaemic myelopoiesis.
      ].
      MDS clonal hematopoiesis is characterized by a myeloid differentiation bias of HSCs, which could be due either to the positive selection of a myeloid-specified MPP or the involution of B-cell progenitors in MDS bone marrow. Sternberg et al. have demonstrated B-cell progenitor depletion in MDS [
      • Sternberg A.
      • Killick S.
      • Littlewood T.
      • et al.
      Evidence for reduced B-cell progenitors in early (low-risk) myelodysplastic syndrome.
      ]. Mutations could accumulate with a higher probability in actively dividing myeloid progenitors that supply the daily production of blood cells than in lymphoid cells.
      It is remarkable that the size of the HSC compartment where the disease initiates remains small. This encourages searching for disease-propagating cells at the level of MPPs. As first evidence for the involvement of MPPs, the engraftment of a clonal linCD34+CD38CD90/Thy1CD45RA MPP into NSG mice has been obtained [
      • Rouault-Pierre K.
      • Smith A.E.
      • Mian S.A.
      • et al.
      Myelodysplastic syndrome can propagate from the multipotent progenitor compartment.
      ]. In this case, the MPP harbored an ASXL1 mutation not detected in HSCs that may confer a proliferative advantage to MPP in vivo [
      • Rouault-Pierre K.
      • Smith A.E.
      • Mian S.A.
      • et al.
      Myelodysplastic syndrome can propagate from the multipotent progenitor compartment.
      ]. In another MDS with excess of blasts >10%, we detected seven mutations in the STAG2, ETV6, U2AF1, ASXL1, and NRAS genes at the level of bulk mononuclear cells, including one STAG2 p.798fs (insC) represented with a variant allele frequency (VAF) over 50% in all compartments and two STAG2 p.A916fs (ins TAAG) and STAG2 p.R1045X detected at 5% and 10% of cells, respectively. These latter mutations were absent in HSCs, detected in MPPs and transmitted with the same VAF to CMPs, GMPs, and MEPs (Fig. 1). This indicates that MDS bone marrow can be made of a variegation of mutations scattered between distinct stem or progenitor subsets. We observed the emergence of three STAG2-mutated subclones on a common genetic background among MPPs. Whether these clones could initiate the disease in NSG mice is not known. In this case, the HSC remains the disease-initiating cell. The concept of several coexisting cancer stem cells developed from the analysis of clear cell renal cancer architecture cannot be applied to MDS [
      • Gerlinger M.
      • Rowan A.J.
      • Horswell S.
      • et al.
      Intratumor heterogeneity and branched evolution revealed by multiregion sequencing.
      ].
      Figure 1
      Figure 1Hierarchy of clonal hematopoiesis. (a) Schematic representation of normal hematopoiesis. (b) Low-risk MDS showing amplification of CMPs and diminution of CLPs. (c) High-risk MDS showing amplification of GMPs. (d) Hierarchy of clonal hematopoiesis in two examples of low-risk MDS by genotype analysis of sorted bone marrow subpopulations. #1: Onset of an “f” mutation in GMP; #2: acquisition of two “f” and “g” mutations in two separate clones in MPPs transmitted to the progeny. Each letter represents one mutation. Gr = granulocyte, Er = erythrocyte, Mo = monocyte, Plt = platelet.
      Considering the diversity of phenotypes, it is not surprising that the hierarchical trees are highly variable in MDS. Will et al. studied the repartition of committed progenitors within the CD34+CD38+ compartment and showed that CMPs accumulate at the expense of MEPs and GMPs in early MDS [
      • Will B.
      • Zhou L.
      • Vogler T.O.
      • et al.
      Stem and progenitor cells in myelodysplastic syndromes show aberrant stage-specific expansion and harbor genetic and epigenetic alterations.
      ]. In these latter cases, CMPs or GMPs may derive from a rapidly proliferating myeloid-specified MPP devoid of erythro-megakaryocytic differentiation capacities. Alternatively, mutations may increase the competitiveness of CMPs or GMPs (Fig. 1). We have shown that a BCOR mutation identified as being dominant in CMPs may become undetectable in GMPs or MEPs. This suggests that the capacity of the BCOR-mutated subclone to differentiate toward the granulocytic and erythroid lineages is abrogated [
      • Chesnais V.
      • Arcangeli M.L.
      • Delette C.
      • et al.
      Architectural and functional heterogeneity of hematopoietic stem/progenitor cells in non-del(5q) myelodysplastic syndromes.
      ]. Invalidation of BCOR using an shRNA strategy in normal progenitors confirmed this hypothesis.
      By using classical biomarkers for CMP LinCD34+CD38+CD123+ CD45RA, GMP LinCD34+CD38+CD123+CD45RA+, and MEP LinCD34+CD38+CD123CD45RA, we showed that the MEPs are well preserved or increased in sideroblastic anemias, whereas this compartment is diminished in other subtypes [
      • Chesnais V.
      • Arcangeli M.L.
      • Delette C.
      • et al.
      Architectural and functional heterogeneity of hematopoietic stem/progenitor cells in non-del(5q) myelodysplastic syndromes.
      ]. MEPs could be fed by selected erythroid-biased MPPs.
      As a hallmark of the myeloid and lymphoid potential of disease-initiating cells, B lymphocytes and, to a lesser extent, T lymphocytes contain MDS-recurrent mutations and participate in clonal hematopoiesis [
      • Woll P.S.
      • Kjällquist U.
      • Chowdhury O.
      • et al.
      Myelodysplastic syndromes are propagated by rare and distinct human cancer stem cells in vivo.
      ,
      • Smith A.E.
      • Mohamedali A.M.
      • Kulasekararaj A.
      • et al.
      Next-generation sequencing of the TET2 gene in 355 MDS and CMML patients reveals low-abundance mutant clones with early origins, but indicates no definite prognostic value.
      ,
      • Medyouf H.
      • Mossner M.
      • Jann J.C.
      • et al.
      Myelodysplastic cells in patients reprogram mesenchymal stromal cells to establish a transplantable stem cell niche disease unit.
      ]. However, the mutational repertoire of B and T lymphocytes is limited. Mutations of the founding clone, mainly affecting epigenetic regulators, splicing factors, and chromatin modifiers are detectable in B lymphocytes [
      • Mossner M.
      • Jann J.C.
      • Wittig J.
      • et al.
      Mutational hierarchies in myelodysplastic syndromes dynamically adapt and evolve upon therapy response and failure.
      ,
      • Chesnais V.
      • Arcangeli M.L.
      • Delette C.
      • et al.
      Architectural and functional heterogeneity of hematopoietic stem/progenitor cells in non-del(5q) myelodysplastic syndromes.
      ]. B lymphocytes are usually less mutated than myeloid cells. Several hypotheses can be put forward: (1) the onset of secondary mutations in the context of initial genetic background may impair the transition of HSCs to lymphoid progenitors and decrease the fitness of B cells; (2) secondary mutations may appear in an MPP with low lymphoid potential [
      • Pietras E.M.
      • Reynaud D.
      • Kang Y.-A.
      • et al.
      Functionally distinct subsets of lineage-biased multipotent progenitors control blood production in normal and regenerative conditions.
      ,
      • Benz C.
      • Copley M.R.
      • Kent D.G.
      • et al.
      Hematopoietic stem cell subtypes expand differentially during development and display distinct lymphopoietic programs.
      ]; and (3) mature B cells may originate from the HSPC compartment downstream of a long-lived lymphoid progenitor affected by early genetic lesions but with slow proliferative capacity that precludes the acquisition of additional events [
      • Sun J.
      • Ramos A.
      • Chapman B.
      • et al.
      Clonal dynamics of native haematopoiesis.
      ,
      • McDermott D.H.
      • Gao J.L.
      • Liu Q.
      • et al.
      Chromothriptic cure of WHIM syndrome.
      ,
      • Busch K.
      • Klapproth K.
      • Barile M.
      • et al.
      Fundamental properties of unperturbed haematopoiesis from stem cells in vivo.
      ]. Differences in mutation composition between myeloid and lymphoid compartments contribute to tumor heterogeneity.
      In the context of differentiation, the various combinations of mutations take part in the selection of myeloid compartments. In other words, the function of targeted genes participates in the selection and the functional consequences of altering one gene depends on the cell type in which the alteration occurs and on associated events [
      • Vogelstein B.
      • Papadopoulos N.
      • Velculescu V.E.
      • Zhou S.
      • Diaz Jr, L.A.
      • Kinzler K.W.
      Cancer genome landscapes.
      ]. The genetic progression and phenotype is linked to the tissue of origin, the context-dependent selection of pro-tumorigenic alterations, and the functional interactions of cooperating genes [
      • de Bruin E.C.
      • McGranahan N.
      • Mitter R.
      • et al.
      Spatial and temporal diversity in genomic instability processes defines lung cancer evolution.
      ,
      • Mina M.
      • Raynaud F.
      • Tavernari D.
      • et al.
      Conditional selection of genomic alterations dictates cancer evolution and oncogenic dependencies.
      ].

      Leukemia-propagating cells

      In AML, clonal hematopoiesis is driven by a leukemic stem cell (LSC) [
      • Lapidot T.
      • Sirard C.
      • Vormoor J.
      • et al.
      A cell initiating human acute myeloid leukaemia after transplantation into SCID mice.
      ,
      • Bonnet D.
      • Dick J.E.
      Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell.
      ]. This cell retains the self-renewal capability of its normal counterpart in NSG mice. The clone that usually engrafts may be a subclone distinct from the one that emerges at relapse and distinct from the putative founding clone [
      • Pearce D.J.
      • Taussig D.
      • Zibara K.
      • et al.
      AML engraftment in the NOD/SCID assay reflects the outcome of AML: implications for our understanding of the heterogeneity of AML.
      ,
      • Klco J.M.
      • Spencer D.H.
      • Miller C.A.
      • et al.
      Functional heterogeneity of genetically defined subclones in acute myeloid leukemia.
      ]. In CD34+ AML, the hierarchical organization of hematopoiesis is highly perturbed. LSC activity has been identified in two hierarchically organized populations distinct from HSC, one being similar to LMPPs and the other resembling normal GMPs. These two populations retain a self-renewal potential [
      • Goardon N.
      • Marchi E.
      • Atzberger A.
      • et al.
      Coexistence of LMPP-like and GMP-like leukemia stem cells in acute myeloid leukemia.
      ].
      In advanced MDS entering the process of AML transformation, the propagating population could be the GMPs because the percentage of GMPs increased [
      • Will B.
      • Zhou L.
      • Vogler T.O.
      • et al.
      Stem and progenitor cells in myelodysplastic syndromes show aberrant stage-specific expansion and harbor genetic and epigenetic alterations.
      ]. We have reported that a STAG2 gene mutation that was faintly detectable in HSCs (LinCD34+CD38CD90+) can amplify in a GMP. Among three distinct clones coexisting in the GMP compartment at the time of diagnosis, the clone containing this mutation amplified as the disease progressed from a MDS with excess of blasts type 1 to a MDS with excess of blasts type 2 and AML at the expense of the two other clones [
      • Chesnais V.
      • Arcangeli M.L.
      • Delette C.
      • et al.
      Architectural and functional heterogeneity of hematopoietic stem/progenitor cells in non-del(5q) myelodysplastic syndromes.
      ]. In the context of lineage commitment, some mutations may increase the fitness of clones to their microenvironment and help the leukemic transformation. As shown in the case of de novo AML, the expansion of a population of cells with putative self-renewal potential downstream of the stem cell compartment should be a key step in the transition from MDS to AML. All initiating mutations are still present as the disease evolves into AML; however, the variant allele frequencies may be different due to the amplification of the leukemic clone.
      Deciphering clonal hematopoiesis during AML initiation and evolution has consolidated the concept of preleukemic events. A few genetic events can fit the definition on the basis of the repopulating advantage in xenograft experiments of a HSC containing such a mutation, its persistence in remission samples, and its involvement in the clone driving the relapse. TET2, DNMT3A, and ASXL1 mutations and MLL rearrangements that may lead to leukemic and nonleukemic lymphoid repopulation are preleukemic lesions that can reinitiate the leukemic process at various stages until relapse [
      • Shlush L.I.
      • Zandi S.
      • Mitchell A.
      • et al.
      Identification of pre-leukaemic haematopoietic stem cells in acute leukaemia.
      ,
      • Hirsch P.
      • Zhang Y.
      • Tang R.
      • et al.
      Genetic hierarchy and temporal variegation in the clonal history of acute myeloid leukaemia.
      ].

      Implications for treatment

      A dynamic evolution that comes with a risk of clonal selection and treatment resistance or escape is the heterogeneity seen in MDS and other myeloid malignancies. We have reported a 10 year follow-up of a refractory anemia with excess blasts 1 (RAEB1) patient that was remarkably stable in terms of bone marrow blast percentage for 9 years and entered RAEB2 after that. This patient did not received disease-modifying therapy for 9 years and successively acquired new mutations in the HSPC compartment. At the time of clinical evolution, the architecture became branched, with two separate clones each containing a new mutation in two distinct signaling adapters (Fig. 2). It is clear that, under treatment with growth factors such as an erythropoiesis-stimulating agent, a recombinant thrombopoietic mimetic, or G-CSF, new mutations still accumulate. Lenalidomide clearly exerts a pressure leading to the disruption of the del(5q) clone and a balanced amplification of non-del(5q) clones. Azacytidine does not prevent the onset of new mutations such as EZH2 [
      • Mossner M.
      • Jann J.C.
      • Wittig J.
      • et al.
      Mutational hierarchies in myelodysplastic syndromes dynamically adapt and evolve upon therapy response and failure.
      ,
      • da Silva-Coelho P.
      • Kroeze L.I.
      • Yoshida K.
      • et al.
      Clonal evolution in myelodysplastic syndromes.
      ].
      Figure 2
      Figure 2Clonal evolution over time. (a) Natural course of clonal evolution: linear acquisition until 90 months (mo) and branched architecture after 119 months of evolution. (b) Clonal evolution under treatment: shown are two examples of clonal evolution under lenalidomide. Upper panel: modification of clone size in one responding patient. Lower panel: acquisition of a new mutation and amplification of the new clone in one nonresponding patient. Each letter represents a mutation.
      In 5q– syndrome or MDS with isolated del(5q), lenalidomide is a powerful drug that rescues patients from anemia, improves quality of life, and prolongs survival [
      • Komrokji R.S.
      • List A.F.
      Short- and long-term benefits of lenalidomide treatment in patients with lower-risk del(5q) myelodysplastic syndromes.
      ]. However, this drug targets 5q– erythroid cells but fails to eradicate the 5q– HSCs responsible for disease maintenance [
      • Tehranchi R.
      • Woll P.S.
      • Anderson K.
      • et al.
      Persistent malignant stem cells in del(5q) myelodysplasia in remission.
      ]. Several scenarios are possible [
      • Mossner M.
      • Jann J.C.
      • Wittig J.
      • et al.
      Mutational hierarchies in myelodysplastic syndromes dynamically adapt and evolve upon therapy response and failure.
      ,
      • da Silva-Coelho P.
      • Kroeze L.I.
      • Yoshida K.
      • et al.
      Clonal evolution in myelodysplastic syndromes.
      ]. In most cases, lenalidomide abrogates the expansion of the founding clone when it contains the del(5q) abnormality, leading or not to the expansion of a separate clone. In non-del(5q) MDS, lenalidomide is less efficient but may act by targeting the hematopoietic cell directly, interfering with stromal cell interaction, or inducing an immune response. We have shown that lenalidomide targets HSPCs transiently by decreasing the size of the clones in responding patients. The loss of response is associated with a re-increase of clone size and sometimes with emergence of new mutations (Fig. 2) [
      • Chesnais V.
      • Renneville A.
      • Toma A.
      • et al.
      Effect of lenalidomide treatment on clonal architecture of myelodysplastic syndromes without 5q deletion.
      ]. Lenalidomide also modulates the immune populations of the bone marrow microenvironment by increasing the abundance of the cytotoxic T-cell population, suggesting that it may also induce an indirect immune-mediated antitumor effect [
      • Kerdivel G.
      • Chesnais V.
      • Becht E.
      • et al.
      Lenalidomide-mediated erythroid improvement in non-del(5q) myelodysplastic syndromes is associated with bone marrow immuno-remodeling.
      ].
      Targeting the HSCs is elusive because these cells are partly dormant and less responsive to drugs than more differentiated cells [
      • Chesnais V.
      • Arcangeli M.L.
      • Delette C.
      • et al.
      Architectural and functional heterogeneity of hematopoietic stem/progenitor cells in non-del(5q) myelodysplastic syndromes.
      ,
      • Tehranchi R.
      • Woll P.S.
      • Anderson K.
      • et al.
      Persistent malignant stem cells in del(5q) myelodysplasia in remission.
      ]. Furthermore, in MDS patients, who are mostly older than 70 years, it is unrealistic to target HSCs because the residual normal HSC is weak and plausibly functionally inefficient to repopulate the bone marrow after treatment. Sparing normal HSCs is therefore challenging, particularly in cases of advanced MDS. The precise identification of disease-propagating cells could be contributive to the design of a therapy specifically directed against them to slow down the progression. Single-cell genetic profiling will provide the best resolution to decipher the clonal heterogeneity and identify the self-renewing compartment in which the disease amplifies. Single-cell DNA screening using a microfluidic platform coupled to multiplex fluorescent quantitative polymerase chain reaction allowing unbiased single-cell selection will provide a comprehensive analysis of genetic abnormalities at diagnosis and during evolution and will also serve in the detection of residual disease. This approach could reveal the potential sources of treatment resistance or failure to guide the choice of alternative therapies.

      Conclusion

      MDS initiates with mutations in epigenetic regulators and/or splice factors at the level of HSCs. Clonal evolution depends on the acquisition of a secondary event usually in the compartment of HSCs, but also in MPPs or GMPs, showing that the context of differentiation may be favorable to the selection of a clone that drives the transformation. Further work is needed to identify specific biomarkers of the disease-propagating compartment that could be expressed in the context of particular patterns of mutated genes. This could help to redirect to this compartment targeted therapies against signaling molecules or deregulated metabolism pathways that appear as frequent secondary transforming events.

      Conflict of interest disclosure

      The authors declare no competing financial interests.

      Acknowledgments

      The authors thank Drs. M.L. Arcangeli, O. Kosmider, E. Lauret, and F. Pflumio for their contribution to this work.
      This work was supported by grants from the Institut National du Cancer ( Recherche Clinique MDS-04 , Grant number: INCa-DGOS_5480 ), Site Intégré de Recherche sur le Cancer CARPEM , and Institut National de la Santé et de la Recherche Médicale (INSERM) .

      References

        • Corey S.J.
        • Minden M.D.
        • Barber D.L.
        • Kantarjian H.
        • Wang J.C.
        • Schimmer A.D.
        Myelodysplastic syndromes: the complexity of stem-cell diseases.
        Nat Rev Cancer. 2007; 7: 118-129
        • Arber D.A.
        • Orazi A.
        • Hasserjian R.
        • et al.
        The 2016 revision to the World Health Organization classification of myeloid neoplasms and acute leukemia.
        Blood. 2016; 127: 2391-2405
        • Delhommeau F.
        • Dupont S.
        • Della Valle V.
        • et al.
        Mutation in TET2 in myeloid cancers.
        N Engl J Med. 2009; 360: 2289-2301
        • Gelsi-Boyer V.
        • Trouplin V.
        • Adélaïde J.
        • et al.
        Mutations of polycomb-associated gene ASXL1 in myelodysplastic syndromes and chronic myelomonocytic leukaemia.
        Br J Haematol. 2009; 145: 788-800
        • Ernst T.
        • Chase A.J.
        • Score J.
        • et al.
        Inactivating mutations of the histone methyltransferase gene EZH2 in myeloid disorders.
        Nat Genet. 2010; 42: 722-726
        • Yoshida K.
        • Sanada M.
        • Shiraishi Y.
        • et al.
        Frequent pathway mutations of splicing machinery in myelodysplasia.
        Nature. 2011; 478: 64-69
        • Quivoron C.
        • Couronné L.
        • Della Valle V.
        • et al.
        TET2 inactivation results in pleiotropic hematopoietic abnormalities in mouse and is a recurrent event during human lymphomagenesis.
        Cancer Cell. 2011; 20: 25-38
        • Wang L.
        • Lawrence M.S.
        • Wan Y.
        • et al.
        SF3B1 and other novel cancer genes in chronic lymphocytic leukemia.
        N Engl J Med. 2011; 365: 2497-2506
        • Alexandrov L.B.
        • Nik-Zainal S.
        • Wedge D.C.
        • et al.
        Signatures of mutational processes in human cancer.
        Nature. 2013; 500: 415-421
        • Pepper J.W.
        • Sprouffske K.
        • Maley C.C.
        Animal cell differentiation patterns suppress somatic evolution.
        PLoS Comput Biol. 2007; 3: e250
        • Tomasetti C.
        • Vogelstein B.
        • Parmigiani G.
        Half or more of the somatic mutations in cancers of self-renewing tissues originate prior to tumor initiation.
        Proc Natl Acad Sci USA. 2013; 110: 1999-2004
        • Walter D.
        • Lier A.
        • Geiselhart A.
        • et al.
        Exit from dormancy provokes DNA-damage-induced attrition in haematopoietic stem cells.
        Nature. 2015; 520: 549-552
        • Busque L.
        • Patel J.P.
        • Figueroa M.E.
        • et al.
        Recurrent somatic TET2 mutations in normal elderly individuals with clonal hematopoiesis.
        Nat Genet. 2012; 44: 1179-1181
        • Xie M.
        • Lu C.
        • Wang J.
        • et al.
        Age-related mutations associated with clonal hematopoietic expansion and malignancies.
        Nat Med. 2014; 20: 1472-1478
        • Jaiswal S.
        • Fontanillas P.
        • Flannick J.
        • et al.
        Age-related clonal hematopoiesis associated with adverse outcomes.
        N Engl J Med. 2014; 371: 2488-2498
        • Genovese G.
        • Kähler A.K.
        • Handsaker R.E.
        • et al.
        Clonal hematopoiesis and blood-cancer risk inferred from blood DNA sequence.
        N Engl J Med. 2014; 371: 2477-2487
        • McKerrell T.
        • Park N.
        • Moreno T.
        • et al.
        Leukemia-associated somatic mutations drive distinct patterns of age-related clonal hematopoiesis.
        Cell Rep. 2015; 10: 1239-1245
        • Steensma D.P.
        • Bejar R.
        • Jaiswal S.
        • et al.
        Clonal hematopoiesis of indeterminate potential and its distinction from myelodysplastic syndromes.
        Blood. 2015; 126: 9-16
        • Cargo C.A.
        • Rowbotham N.
        • Evans P.A.
        • et al.
        Targeted sequencing identifies patients with preclinical MDS at high risk of disease progression.
        Blood. 2015; 126: 2362-2365
        • Walter M.J.
        • Shen D.
        • Ding L.
        • et al.
        Clonal architecture of secondary acute myeloid leukemia.
        N Engl J Med. 2012; 366: 1090-1098
        • Mossner M.
        • Jann J.C.
        • Wittig J.
        • et al.
        Mutational hierarchies in myelodysplastic syndromes dynamically adapt and evolve upon therapy response and failure.
        Blood. 2016; 128: 1246-1259
        • da Silva-Coelho P.
        • Kroeze L.I.
        • Yoshida K.
        • et al.
        Clonal evolution in myelodysplastic syndromes.
        Nat Commun. 2017; 8: 15099
        • Damm F.
        • Chesnais V.
        • Nagata Y.
        • et al.
        BCOR and BCORL1 mutations in myelodysplastic syndromes and related disorders.
        Blood. 2013; 122: 3169-3177
        • Mazumdar C.
        • Shen Y.
        • Xavy S.
        • et al.
        Leukemia-associated cohesin mutants dominantly enforce stem cell programs and impair human hematopoietic progenitor differentiation.
        Cell Stem Cell. 2015; 17: 675-688
        • Malcovati L.
        • Karimi M.
        • Papaemmanuil E.
        • et al.
        SF3B1 mutation identifies a distinct subset of myelodysplastic syndrome with ring sideroblasts.
        Blood. 2015; 126: 233-241
        • Ortmann C.A.
        • Kent D.G.
        • Nangalia J.
        • et al.
        Effect of mutation order on myeloproliferative neoplasms.
        N Engl J Med. 2015; 372: 601-612
        • Chesnais V.
        • Arcangeli M.L.
        • Delette C.
        • et al.
        Architectural and functional heterogeneity of hematopoietic stem/progenitor cells in non-del(5q) myelodysplastic syndromes.
        Blood. 2017; 129: 484-496
        • Obeng E.A.
        • Chappell R.J.
        • Seiler M.
        • et al.
        Physiologic expression of sf3b1(k700e) causes impaired erythropoiesis, aberrant splicing, and sensitivity to therapeutic spliceosome modulation.
        Cancer Cell. 2016; 30: 404-417
        • Mortera-Blanco T.
        • Dimitriou M.
        • Woll P.S.
        • et al.
        SF3B1-initiating mutations in MDS-RSs target lymphomyeloid hematopoietic stem cells.
        Blood. 2017; 130: 881-890
        • Majeti R.
        • Park C.Y.
        • Weissman I.L.
        Identification of a hierarchy of multipotent hematopoietic progenitors in human cord blood.
        Cell Stem Cell. 2007; 1: 635-645
        • Mazurier F.
        • Doedens M.
        • Gan O.I.
        • Dick J.E.
        Rapid myeloerythroid repopulation after intrafemoral transplantation of NOD-SCID mice reveals a new class of human stem cells.
        Nat Med. 2003; 9: 959-963
        • Dykstra B.
        • Kent D.
        • Bowie M.
        • et al.
        Long-term propagation of distinct hematopoietic differentiation programs in vivo.
        Cell Stem Cell. 2007; 1: 218-229
        • Yamamoto R.
        • Morita Y.
        • Ooehara J.
        • et al.
        Clonal analysis unveils self-renewing lineage-restricted progenitors generated directly from hematopoietic stem cells.
        Cell. 2013; 154: 1112-1126
        • Cabezas-Wallscheid N.
        • Klimmeck D.
        • Hansson J.
        • et al.
        Identification of regulatory networks in HSCs and their immediate progeny via integrated proteome, transcriptome, and DNA methylome analysis.
        Cell Stem Cell. 2014; 15: 507-522
        • Paul F.
        • Arkin Y.
        • Giladi A.
        • et al.
        Transcriptional heterogeneity and lineage commitment in myeloid progenitors.
        Cell. 2015; 163: 1663-1677
        • Perié L.
        • Duffy K.R.
        • Kok L.
        • de Boer R.J.
        • Schumacher T.N.
        The branching point in erythro-myeloid differentiation.
        Cell. 2015; 163: 1655-1662
        • Notta F.
        • Zandi S.
        • Takayama N.
        • et al.
        Distinct routes of lineage development reshape the human blood hierarchy across ontogeny.
        Science. 2016; 351: aab2116
        • Naik S.H.
        • Perié L.
        • Swart E.
        • et al.
        Diverse and heritable lineage imprinting of early haematopoietic progenitors.
        Nature. 2013; 496: 229-232
        • Doulatov S.
        • Notta F.
        • Eppert K.
        • Nguyen L.T.
        • Ohashi P.S.
        • Dick J.E.
        Revised map of the human progenitor hierarchy shows the origin of macrophages and dendritic cells in early lymphoid development.
        Nat Immunol. 2010; 11: 585-593
        • Laurenti E.
        • Doulatov S.
        • Zandi S.
        • et al.
        The transcriptional architecture of early human hematopoiesis identifies multilevel control of lymphoid commitment.
        Nat Immunol. 2013; 14: 756-763
        • Yu V.W.
        • Yusuf R.Z.
        • Oki T.
        • et al.
        Epigenetic memory underlies cell-autonomous heterogeneous behavior of hematopoietic stem cells.
        Cell. 2016; 167: 1310-1322
        • Pietras E.M.
        • Reynaud D.
        • Kang Y.-A.
        • et al.
        Functionally distinct subsets of lineage-biased multipotent progenitors control blood production in normal and regenerative conditions.
        Cell Stem Cell. 2015; 17: 35-46
        • Hoppe P.S.
        • Schwarzfischer M.
        • Loeffler D.
        • et al.
        Early myeloid lineage choice is not initiated by random PU.1 to GATA1 protein ratios.
        Nature. 2016; 535: 299-302
        • Sun J.
        • Ramos A.
        • Chapman B.
        • et al.
        Clonal dynamics of native haematopoiesis.
        Nature. 2014; 514: 322-327
        • Hérault A.
        • Binnewies M.
        • Leong S.
        • et al.
        Myeloid progenitor cluster formation drives emergency and leukaemic myelopoiesis.
        Nature. 2017; 544: 53-58
        • Yahata T.
        • Takanashi T.
        • Muguruma Y.
        • et al.
        Accumulation of oxidative DNA damage restricts the self-renewal capacity of human hematopoietic stem cells.
        Blood. 2011; 118: 2941-2950
        • Beerman I.
        • Bhattacharya D.
        • Zandi S.
        • et al.
        Functionally distinct hematopoietic stem cells modulate hematopoietic lineage potential during aging by a mechanism of clonal expansion.
        Proc Natl Acad Sci USA. 2010; 107: 5465-5470
        • Pang W.W.
        • Price E.A.
        • Sahoo D.
        • et al.
        Human bone marrow hematopoietic stem cells are increased in frequency and myeloid-biased with age.
        Proc Natl Acad Sci USA. 2011; 108: 20012-20017
        • Rossi D.J.
        • Bryder D.
        • Seita J.
        • Nussenzweig A.
        • Hoeijmakers J.
        • Weissman I.L.
        Deficiencies in DNA damage repair limit the function of haematopoietic stem cells with age.
        Nature. 2007; 447: 725-729
        • Beerman I.
        • Seita J.
        • Inlay M.A.
        • Weissman I.L.
        • Rossi D.J.
        Quiescent hematopoietic stem cells accumulate DNA damage during aging that is repaired upon entry into cell cycle.
        Cell Stem Cell. 2014; 15: 37-50
        • Flach J.
        • Bakker S.T.
        • Mohrin M.
        • et al.
        Replication stress is a potent driver of functional decline in ageing haematopoietic stem cells.
        Nature. 2014; 512: 198-202
        • Zambetti N.A.
        • Ping Z.
        • Chen S.
        • et al.
        Mesenchymal inflammation drives genotoxic stress in hematopoietic stem cells and predicts disease evolution in human pre-leukemia.
        Cell Stem Cell. 2016; 19: 613-627
        • Schneider R.K.
        • Schenone M.
        • Ferreira M.V.
        • et al.
        Rps14 haploinsufficiency causes a block in erythroid differentiation mediated by S100A8 and S100A9.
        Nat Med. 2016; 22: 288-297
        • Nilsson L.
        • Astrand-Grundström I.
        • Arvidsson I.
        • et al.
        Isolation and characterization of hematopoietic progenitor/stem cells in 5q-deleted myelodysplastic syndromes: evidence for involvement at the hematopoietic stem cell level.
        Blood. 2000; 96: 2012-2021
        • Tehranchi R.
        • Woll P.S.
        • Anderson K.
        • et al.
        Persistent malignant stem cells in del(5q) myelodysplasia in remission.
        N Engl J Med. 2010; 363: 1025-1037
        • Woll P.S.
        • Kjällquist U.
        • Chowdhury O.
        • et al.
        Myelodysplastic syndromes are propagated by rare and distinct human cancer stem cells in vivo.
        Cancer Cell. 2014; 25: 794-808
        • Thanopoulou E.
        • Cashman J.
        • Kakagianne T.
        • Eaves A.
        • Zoumbos N.
        • Eaves C.
        Engraftment of NOD/SCID-beta2 microglobulin null mice with multilineage neoplastic cells from patients with myelodysplastic syndrome.
        Blood. 2004; 103: 4285-4293
        • Mian S.A.
        • Rouault-Pierre K.
        • Smith A.E.
        • et al.
        SF3B1 mutant MDS-initiating cells may arise from the haematopoietic stem cell compartment.
        Nat Commun. 2015; 6: 10004
        • Wang J.C.
        • Lapidot T.
        • Cashman J.D.
        • et al.
        High level engraftment of NOD/SCID mice by primitive normal and leukemic hematopoietic cells from patients with chronic myeloid leukemia in chronic phase.
        Blood. 1998; 91: 2406-2414
        • Corbin A.S.
        • Agarwal A.
        • Loriaux M.
        • Cortes J.
        • Deininger M.W.
        • Druker B.J.
        Human chronic myeloid leukemia stem cells are insensitive to imatinib despite inhibition of BCR-ABL activity.
        J Clin Invest. 2011; 121: 396-409
        • Sternberg A.
        • Killick S.
        • Littlewood T.
        • et al.
        Evidence for reduced B-cell progenitors in early (low-risk) myelodysplastic syndrome.
        Blood. 2005; 106: 2982-2991
        • Rouault-Pierre K.
        • Smith A.E.
        • Mian S.A.
        • et al.
        Myelodysplastic syndrome can propagate from the multipotent progenitor compartment.
        Haematologica. 2017; 102: e7-e10
        • Gerlinger M.
        • Rowan A.J.
        • Horswell S.
        • et al.
        Intratumor heterogeneity and branched evolution revealed by multiregion sequencing.
        N Engl J Med. 2012; 366: 883-892
        • Will B.
        • Zhou L.
        • Vogler T.O.
        • et al.
        Stem and progenitor cells in myelodysplastic syndromes show aberrant stage-specific expansion and harbor genetic and epigenetic alterations.
        Blood. 2012; 120: 2076-2086
        • Smith A.E.
        • Mohamedali A.M.
        • Kulasekararaj A.
        • et al.
        Next-generation sequencing of the TET2 gene in 355 MDS and CMML patients reveals low-abundance mutant clones with early origins, but indicates no definite prognostic value.
        Blood. 2010; 116: 3923-3932
        • Medyouf H.
        • Mossner M.
        • Jann J.C.
        • et al.
        Myelodysplastic cells in patients reprogram mesenchymal stromal cells to establish a transplantable stem cell niche disease unit.
        Cell Stem Cell. 2014; 14: 824-837
        • Benz C.
        • Copley M.R.
        • Kent D.G.
        • et al.
        Hematopoietic stem cell subtypes expand differentially during development and display distinct lymphopoietic programs.
        Cell Stem Cell. 2012; 10: 273-283
        • McDermott D.H.
        • Gao J.L.
        • Liu Q.
        • et al.
        Chromothriptic cure of WHIM syndrome.
        Cell. 2015; 160: 686-699
        • Busch K.
        • Klapproth K.
        • Barile M.
        • et al.
        Fundamental properties of unperturbed haematopoiesis from stem cells in vivo.
        Nature. 2015; 518: 542-546
        • Vogelstein B.
        • Papadopoulos N.
        • Velculescu V.E.
        • Zhou S.
        • Diaz Jr, L.A.
        • Kinzler K.W.
        Cancer genome landscapes.
        Science. 2013; 339: 1546-1558
        • de Bruin E.C.
        • McGranahan N.
        • Mitter R.
        • et al.
        Spatial and temporal diversity in genomic instability processes defines lung cancer evolution.
        Science. 2014; 346: 251-256
        • Mina M.
        • Raynaud F.
        • Tavernari D.
        • et al.
        Conditional selection of genomic alterations dictates cancer evolution and oncogenic dependencies.
        Cancer Cell. 2017; 32: 155-168
        • Lapidot T.
        • Sirard C.
        • Vormoor J.
        • et al.
        A cell initiating human acute myeloid leukaemia after transplantation into SCID mice.
        Nature. 1994; 367: 645-648
        • Bonnet D.
        • Dick J.E.
        Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell.
        Nat Med. 1997; 3: 730-737
        • Pearce D.J.
        • Taussig D.
        • Zibara K.
        • et al.
        AML engraftment in the NOD/SCID assay reflects the outcome of AML: implications for our understanding of the heterogeneity of AML.
        Blood. 2006; 107: 1166-1173
        • Klco J.M.
        • Spencer D.H.
        • Miller C.A.
        • et al.
        Functional heterogeneity of genetically defined subclones in acute myeloid leukemia.
        Cancer Cell. 2014; 25: 379-392
        • Goardon N.
        • Marchi E.
        • Atzberger A.
        • et al.
        Coexistence of LMPP-like and GMP-like leukemia stem cells in acute myeloid leukemia.
        Cancer Cell. 2011; 19: 138-152
        • Shlush L.I.
        • Zandi S.
        • Mitchell A.
        • et al.
        Identification of pre-leukaemic haematopoietic stem cells in acute leukaemia.
        Nature. 2014; 506: 328-333
        • Hirsch P.
        • Zhang Y.
        • Tang R.
        • et al.
        Genetic hierarchy and temporal variegation in the clonal history of acute myeloid leukaemia.
        Nat Commun. 2016; 7: 12475
        • Komrokji R.S.
        • List A.F.
        Short- and long-term benefits of lenalidomide treatment in patients with lower-risk del(5q) myelodysplastic syndromes.
        Ann Oncol. 2016; 27: 62-68
        • Chesnais V.
        • Renneville A.
        • Toma A.
        • et al.
        Effect of lenalidomide treatment on clonal architecture of myelodysplastic syndromes without 5q deletion.
        Blood. 2016; 127: 749-760
        • Kerdivel G.
        • Chesnais V.
        • Becht E.
        • et al.
        Lenalidomide-mediated erythroid improvement in non-del(5q) myelodysplastic syndromes is associated with bone marrow immuno-remodeling.
        Leukemia. 2017; (In press)https://doi.org/10.1038/leu.2017.305