JAK2^V617F/TET2 mutations: does the order matter?

According to the World Health Organization classification, myeloproliferative neoplasms (MPN) include chronic myelogenous leukemia, also known as BCR-ABL1–positive MPN, classic BCR-ABL1-negative MPN including polycythemia vera (PV), essential thrombocythemia (ET) and primary myelofibrosis (PMF), and non-classic forms (i.e. systemic mastocytosis, chronic eosinophilic leukemia not otherwise specified, chronic neutrophilic leukemia and unclassifiable MPN). All these subtypes are stem cell-derived clonal myeloproliferation, associated with the overproduction of mature blood elements and variable rates of transformation to acute myeloid leukemia (AML).¹

JAK2^V617F activating mutation is the most prevalent abnormality observed in BCR-ABL1-negative MPN, found in virtually all cases of PV and in about half of ET and PMF (96%, 55% and 65%, respectively). This mutation lies in the pseudokinase-domain of JAK2 and disrupts its regulatory activity. Another mutation affecting JAK2 exon 12 is observed in 3% of all PV cases. Mutations affecting W515 of the thrombopoietin receptor MPL are detected in PMF and ET patients. Additional mutations have been identified in MPN (reviewed in ¹). Defects in the control of intracellular signaling involve mutations in LNK and CBL genes. Genetic abnormalities affecting epigenetic regulation, and possibly responsible for disease initiation, concern the ASXL1, EZH2 and TET2 genes. Finally, mutation in IKZF1 and IDH1/2 may be implicated in MPN transformation. In PV, among these additional mutations to JAK2^V617F, TET2 mutations are those most frequently reported (16%); the others are only described in small subsets of patients.¹

TET2 belongs to a family of three conserved genes in mammals: TET1, TET2 and TET3. The founding member of the family, TET1, has been identified as a fusion partner of MLL in the t(10;11)(q22;q23) translocation of acute leukemia.^2,3 The involvement of TET3 in hematologic disorder has not yet been described. The TET proteins are members of the 2-oxoglutarate (2-OG)- and Fe(II)-dependent dioxygenase that are able to convert 5-methyl-cytosine (5-mC) to 5-hydroxymethylcytosine (hmC).^4,5 Recent reports indicate an important role for TET1 and TET2 (and, therefore, hmC) in the control of ES cell self-renewal and differentiation.⁶ TET3 might be involved in genome reprogramming following fecundation.⁷

5-hydroxymethylation: a novel major player in the epigenetic field

For decades, the implications and impact of 5-mC in human genome has been extensively studied and it is known to be associated with low gene expression. In contrast, little is known about the recently identified hmC. Indeed, the first study reporting a hydroxylated form of 5-mC in mammalian DNA was described in the early 70s.⁸ However, this modified base did not receive full attention until 2 reports demonstrating that hmC accounts for 0.6% and 0.03% of the total nucleotides in Purkinje cells⁹ and murine ES cells,⁵ respectively.

The function of hmC is not yet clear. Several reports have indicated that hmC prevent the binding of proteins interacting with 5-mC and hmC and may represent a step toward demethylation.⁸ Using a novel chemical method, Song et al.¹⁰ showed an enrichment of hmC in maturing murine brain cells that increases with age and is associated with gene expression. Contrasting results have been reported in human samples: TET2 mutation or inhibition of its catalytic activity affect the level of hmC in myeloid malignant samples, but was associated with a decrease of 5-mC in MDS (myelodysplastic syndromes) samples and with an increase of DNA methylation in AML.^11,12

TET2 function in normal hematopoiesis and MPN

TET2 is expressed in a wide range of tissues, such as kidney, brain and the hematopoietic system.¹³ Recent analyses using short-hairpin RNA in murine stem cells from bone marrow have shown that depletion of TET2 promoted an increase in the proportion of hematopoietic stem/progenitor cells with inhibition of normal differentiation of granulocytes and monocytes.¹² Using a similar approach, TET2 mRNA depletion was also shown to lead to the expansion of monocyte/macrophage cells in the presence of cytokines stimulating granulocytic differentiation (G-CSF or GM-CSF) but not in the presence of macrophage-colony stimulating factor (M-CSF).¹¹ This study indicates that TET2 may act at many different phases of myeloid differentiation.

Acquired inactivating mutations of the TET2 gene are frequently observed in human myeloid malignancies (i.e. MDS, de novo and secondary AML) suggesting that they affect a very early myeloid progenitor. TET2 mutations are found in subsets of all subtypes of MPN (Table 1). Inherited TET2 mutations are not responsible for familial MPN,¹⁴ but a germline mutation of TET2 was recently described in 2 sisters, one suffering from a JAK2^V617F-positive PV, the other healthy.¹⁵

Table 1.

TET2 mutations in BCR-ABL1-negative myeloproliferative neoplasms.

In this issue, Swierczek and colleagues²¹ have investigated the clonality and allele burdens of JAK2^V617F and TET2 mutations in patients with sporadic or familial PV. In contrast to other PV patients described in the initial study,¹⁷ and in agreement with recent reports,¹⁵ they found evidence that TET2 mutations may follow rather than precede the JAK2^V617F mutation in some PV patients. Therefore, no strict temporal order of appearance applies to TET2 and JAK2 mutations in PV. Furthermore, they show that clonal in vitro amplifications of mutated erythroid cells only occur from patient samples having both mutations, suggesting that the presence of TET2 mutations increases the aggression of the PV clone. The questions addressed by this work come down to the respective roles of TET2 and JAK2 mutations in the development of human PV.

Animal models using retroviral delivery, transgenic or knock-in technology have proved conclusively that the sole JAK2^V617F gene expression in hematopoietic cells is able to induce most of the PV phenotype as soon as the levels match at least that of the JAK2^WT allele (heterozygous status). However, does it mean that JAK2^V617F is self sufficient for PV in humans? These mouse models are polyclonal and one can argue that JAK2^V617F is sufficient for PV as soon as there are enough mutated stem cells. Yet in KI models, the ability of JAK2^V617F to amplify the stem cell is controversial.²² Should an additional defect occur for clonal dominance? In human sporadic MPN, other genetic events preceding JAK2^V617F are suspected from clonal analysis, using the X chromosome activation or the 20qdel chromosomal abnormality,¹ and secondary JAK2^WT AML progressing from JAK2^V617F MPN.^23,24 Indeed, TET2 mutations might be instrumental in amplifying immature myeloid cells, as suggested by several reports on human samples.^17,23 However, is it necessary for PV occurrence? The absence of TET2 mutations is reported in most PV patients and in vitro or in vivo mouse SCID clonal amplification seems to be missing in those patients lacking TET2 mutations, as suggested by Swierczek and colleagues.^17,21

In this context, TET2 mutations following or preceding JAK2^V617F could, therefore, be seen as paradigmatic of events possibly predisposing to PV occurrence and/or accelerating PV evolution. Establishing the importance of JAK2^V617F and other TET2-type of mutation orders, their association with transformation and prognosis value, and the precise contribution of each mutation in PV occurrence/evolution will need extensive, in depth clinical, cellular and molecular analyses of a large number of human samples, and animal models demonstrating the precise effect of the identified mutation(s) in a controlled genetic background and their cooperation with JAK2^V617F.