From 1 to 2 mutated JAK2: the road from ET to PV

Comment on Scott et al, page 2435

In this issue of Blood, Scott and colleagues give insight into myeloproliferative disorder heterogeneity induced by V617F JAK2 by showing that ET and PV are associated with mono- and biallelic mutations, respectively.

The discovery in 2005 by 4 different groups that a single JAK2 mutation (V617F JAK2) was observed in 90% of polycythemia vera (PV), 60% of essential thrombocythemia (ET), and 50% of idiopathic myelofibrosis represents a major advance in the understanding of these disorders (and validated the previous theory of the proximity of these myeloproliferative disorders),^1-4  but it has also raised a new major question: how can a single mutation give rise to several diseases? The paper by Scott and colleagues, from Green's group, in this issue of Blood represents one more step to validate the concept that the number of V617F JAK2 copies, as well as the overall intensity of the constitutive signaling, explains the phenotype heterogeneity. It has already been shown that about one third of patients with PV have a biallelic (homozygous) V617F JAK2 mutation associated with a 9p loss of heterozygosity (LOH).⁵  This biallelic mutation appears to be due to a mitotic recombination. In contrast, this is extremely rare in ET. In this report, Scott and colleagues have taken a new step forward by demonstrating that some V617F JAK2 homozygous hematopoietic progenitors were present in all studied patients with PV and undetectable in patients with ET. Thus, this recent work from Green's team strongly suggests that differences between ET and PV are dependent only on this parameter.

V617F JAK2 is a subtle mutation that induces a low gain of function in JAK2. There is increasing evidence that its activity requires the presence of cytokine receptors to induce signaling.⁶  In addition, we reported that mutated V617F JAK2 function was inhibited by the normal nonmutated JAK2.²  Thus, duplication of the mutant allele with the loss of the normal allele will theoretically greatly increase signaling. However, the differences in phenotype between a biallelic and a monoallelic mutation suggest strongly that a biallelic mutation will essentially favor erythroid and granulocytic progenitor proliferation but will inhibit megakaryocytic differentiation. In contrast, a monoallelic mutationwillgiveanadvantagetothemegakaryocytic lineage. This hypothesis explaining disease heterogeneity may underscore major differences in signaling through cytokine receptors and thus open new pathways for research.

Overall, this result suggests that a V617F JAK2 as a single genetic event will give rise to ET. A second genetic event, apparently a mitotic recombination, would be required to cause PV. This hypothesis will absolutely require knock-in mice to be validated. However, it seems likely that any other secondary genetic events synergizing with V617F JAK2 will also be implicated in PV pathogenesis. Thus, it is expected that some patients with PV will be also heterozygous when a larger cohort of patients will be studied. What about idiopathic myelofibrosis: is it related to the same mechanism (ie, the number of V617F JAK2 alleles)? Or are other events necessary to develop fibrosis? Another interesting point concerns the high frequency of mitotic recombination in PV. Does it mean that another event (previous to JAK2 mutation) is necessary to induce a genetic instability before the occurrence of a monoallelic mutation in PV? Indeed, it is expected that if the mono- to biallelic theory is true, all instances of PV will be preceded by ET. However, less than 15% of ETs transform into PV. Are differences between ET and PV thus related to genetic determinants regulating the ability to perform mitotic recombination in the hematopoietic stem cell compartment?

In conclusion, this elegant work from Green's group is a major step in the demonstration that “copy number” is the basis of the heterogeneity of the myeloproliferative disorders, which can explain differences between V617F JAK2 ET and PV, at least in part. ▪