Primary Myelofibrosis (PMF) is a myeloproliferative neoplasm characterized by abnormal differentiation of erythroid-megakaryocytic lineages and expansion of the granulo/monocytic lineage. Accumulation of aberrant myeloid precursors dominates the chronic phase of PMF leading to fibrosis development or leukemic transformation. Recent reports describe that mutation order dictates the prevalence of distinct erythroid subclones in MPN, or that clonality of whole blood mononuclear cells is related to worse prognosis and leukemic transformation. The mutational variability of the stem cell pool determining either the expansion of independent clones dominating chronic phase PMF or the propagation of pre-leukemic progenitors has not been resolved.

In our previous studies, we characterized a CD133+ HSC population exhibiting multilineage differentiation capacity in vitro that drives PMF disease and leukemic transformation in a xenotransplantation mouse model. Molecular analysis of PMF-patient derived HSC indicated variability in their mutational burden, which was reflected in their engraftment capacity and disease induction in vivo. Our goal is to determine the genetic lesions within the HSC pool in PMF that determine aberrant myeloid differentiation in the chronic phase or are responsible for blast transformation.

CD133+ HSCs from 15 PMF patients were molecularly characterized for the known mutations in MPN by whole exon sequencing. Sorted HSC cells were functionally analyzed at a single cell level for variable myeloid colony formation. 2230 colonies were phenotypically characterized and isolated. Analysis of the PMF HSC clonogenic potential indicates that the presence of mutations in the epigenetic regulator EZH2 correlates with granulo/monocytic differentiation but limited erythroid colony formation potential (0-0,05%), as determined in three different patient samples (2 JAK2-V617F+, 1 CALR-fs*+). Transplantation of these patient samples gave the highest engraftment in our mouse model and in one case, EZH2mu JAK2wt leukemic transformation.

CD133+ HSC-derived single colony analysis from this patient indicated that there are 6 different genotypic clones of HSC, which exhibit variable granulo/monocytic differentiation capacity in vitro. From a total of 569 formed colonies, 538 were CFU-GM,-G,-M and 31 BFU-E. PCR analysis of colonies for JAK2-V617F and Sanger sequencing for EZH2-D265H indicates that the presence of JAK2-V617F in hetero- or homozygosity can occur in the EZH2-D265H background without influencing the granulo/monocytic commitment of these mutated HSCs. Interestingly, the limited BFU-Es that arose contained only single JAK2-V617F mutations in the same patient. Moreover, the presence of single EZH2-D265H heterozygous clones, single JAK2-V617F hetero- or homozygous clones, as well as double mutated clones indicate two independent mutational events affecting the same locus and nucleotide have occurred in this patient. In view of the overall high frequency of JAK2-V617F mutations, we predict that the EZH2 mutation was the first mutation in double mutant clones in this patient. Taken together, we show for the first time that JAK2-V617F mutation can occur in independent HSC clones in PMF that exhibit distinctive differentiation potential. Taken into account that AML occurred in vivo from a EZH2mu JAK2wt clone, our studies indicate that JAK2 and CALR mutations sustain the progeny of the chronic phase PMF, while EZH2 mutations might precede those of JAK2 and shape the genomic landscape that supports the expansion of pre-leukemic clones.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution