Different involvement of the megakaryocytic lineage by the JAK2V617F mutation in Polycythemia vera, essential thrombocythemia and chronic idiopathic myelofibrosis
- First Online:
- Cite this article as:
- Hussein, K., Brakensiek, K., Buesche, G. et al. Ann Hematol (2007) 86: 245. doi:10.1007/s00277-007-0252-3
- 96 Views
Atypical megakaryocytes provide the histomorphological hallmark of all Philadelphia-chromosome negative chronic myeloproliferative disorder (Ph− CMPD) subtypes and have not been studied so far for the JAK2V617F mutation. The mutant gene dosage was determined in isolated megakaryocytes from 68 cases of JAK2+/Ph− CMPD by a pyrosequencing assay. Megakaryocytes from essential thrombocythemia (ET) showed significantly lower levels of mutated JAK2 alleles compared to patients with chronic idiopathic myelofibrosis (cIMF) with manifest fibrosis and polycythemia vera (PV) but not to prefibrotic cIMF. Solely, ET JAK2V617F in megakaryocytes is associated with a PV-like phenotype, and at least in one patient, the JAK2 mutation was exclusively acquired within the megakaryocytic lineage. The overt differences between prefibrotic and fibrotic cIMF suggested a causative role of the gene dosage of mutant JAK2 in fibrotic progression. Megakaryocyte analysis of a follow-up of eight individual cases with sequential biopsies, however, showed that progression to homozygosity of V617F mutated JAK2 and onset of manifest fibrosis appeared to be independent events. We conclude that megakaryocytes might be the predominant or even the exclusive lineage that acquires the JAK2V617F mutation in ET and that the JAK2V617F evolution to higher gene dosages represents a dynamic and complex process substantially involving megakaryocytes.
KeywordsMegakaryocytesLaser microdissectionJAK2 V617F mutationPhiladelphia-chromosome negative chronic myeloproliferative disorders
The discovery of the somatic single point mutation (V617F) in the JH2 pseudokinase domain of JAK2 (JAK2V617F) in the majority of patients with polycythemia vera (PV) and in up to 50% of patients with chronic idiopathic myelofibrosis (cIMF) and essential thrombocythemia (ET) provided new insights into the molecular mechanisms that govern the clonal growth of neoplastic stem cells in chronic myeloproliferative disorders (CMPDs) [1–4]. Clonal evolution based on mitotic recombination may occur and lead to a progression from a heterozygous state to homozygosity. There are indications that a correlation between the mutation status and prognosis may exist. Recent clinical studies suggested that patients with cIMF showing mutated JAK2 presented with higher hemoglobin levels, higher white blood cell counts, lower platelets, higher peripheral CD34+ progenitor cell counts, and with a poorer survival compared to patients with wild-type JAK2 . Whether a biallelic JAK2 mutation plays any role in the progression to bone marrow fibrosis has not yet been fully clarified.
In almost all studies on JAK2, purified peripheral cell fractions or alternatively unseparated cells from the bone marrow aspirate were studied but not in bone marrow-derived megakaryocytes because these cells are frequently inaspirable in manifest myelofibrosis (MF), particularly in cIMF. Proliferation of atypical, variably clustered megakaryocytes represents a histopathological key feature in the bone marrow from Philadelphia-chromosome negative (Ph−) CMPDs , and megakaryocyte-derived fibrogenic cytokines are held responsible for the development of MF [7, 8]. Because of their striking atypia and their putative role in pathogenesis of bone marrow fibrosis, isolated megakaryocytes from prefibrotic and advanced cIMF, PV, and ET were taken into focus in this study, and allelic involvement by the V617F mutation of the JAK2 gene was investigated and related to the degree of bone marrow fibrosis.
Materials and methods
Bone marrow study group
Formalin-fixed and paraffin-embedded (FFPE) bone marrow trephine biopsies showing cIMF, PV, and ET were retrieved from the bone marrow registry of the Institute of Pathology, Medizinische Hochschule Hannover, and were re-evaluated according to the World Health Organization classification in close agreement with clinical data and presentation. The study group for determination of mutation frequencies in Ph− CMPD (n = 258, diagnosed between 2000 and 2006) comprised PV (n = 55), prefibrotic cIMF (n = 94), advanced cIMF with manifest MF (n = 71), ET (n = 38), and 63 control cases. The study group was approved independently by three hematopathologists (Buesche, Buhr, Kreipe). The cIMF cases were subdivided into two groups depending on the degree of MF (MF0/1 and MF2/3) as described [9, 10]. From the JAK2V617F-mutated cases (see “Results”), 68 samples were randomly selected for megakaryocyte-specific mutational status (15 prefibrotic cIMF, 15 advanced cIMF, 14 PV, 24 ET). Another seven cIMF cases showing wild-type JAK2 and five cases with megakaryocytic hyperplasia were selected as the control group. In addition, sequential bone marrow trephines representing the disease course of eight patients (range: 1–14 years) with cIMF were investigated for the mutational status of JAK2.
Laser-assisted microdissection for isolation of megakaryocytes
DNA extraction and control PCR
DNA was extracted from total bone marrow cells by Proteinase K digestion of one ∼10-μm-thick section from the FFPE block as described . Genomic DNA was isolatedby using the DNeasy Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions including an obligatory RNA digest using RNase A (AppliChem, Darmstadt, Germany). Microdissected megakaryocytes were directly digested using 100 μl Proteinase K buffer after being catapulted into the lid of a microfuge tube. Control of sufficient DNA extraction was achieved by amplification of a 90 bp β-globin product (B-Glob forward 5′-AGAAGAGCCAAGGACAGGTACG -3′ and B-Glob reverse 5′-TGCTCCTGGGAGTAGATTGGC -3′; GenBank AY260740) from 25 ng DNA.
JAK2 amplification and restriction site analysis
Polymerase chain reaction (PCR) amplification of JAK2 was performed for 40 cycles in a GeneAmp® PCR System 2700 (Applied Biosystems, Weiterstadt, Germany) using 2.5 mM magnesium chloride and 25 ng DNA. Primers for JAK2 amplification covering the hotspot mutation site corresponding to nucleotide 55061 (see GenBank sequence AL161450) were JAK2 forward 5′-TATGATGAGCAAGCTTTCTCACAAG-3′ and JAK2 reverse 5′-AGAAAGGCATTAGAAAGCCTGTAGTT-3′, generating a 102-bp product.
Ten microliters of a given sample were incubated with four units of the restriction enzyme BsaX I (2 U/μl, New England Biolabs, Beverly, MA) for 6 h at 37°C in a PCR cycler with a heated lid to avoid sample evaporation. The BsaX I enzyme recognized a site (5′.... 9(N) A C (N)5 C T C C (N)10 ....3′) that included the hotspot nucleotide 55061 (GenBank sequence AL161450). For determination of BsaX I cuts in the respective JAK2 sequence, the NEBcutter 2.0 software tool was used (http://tools.neb.com/NEBcutter2/index.php). Samples analyzed by restriction site analysis were strictly accompanied by a positive sample (JAK2V617F mutated cell line HEL) and a JAK2 wild-type sample (cell line HL-60 or normal hematopoiesis).
JAK2V617F mutant allele quantification by pyrosequencing®
We established a pyrosequencing assay to quantify the frequency of mutant T alleles in megakaryocytes isolated from Ph− CMPD and disease courses with known JAK2 mutation. For JAK2 amplification, the identical primer sequences and control cell lines (HEL, HL-60) were used as mentioned above. PCR reactions containing 400 nM of forward and biotinylated reverse PCR primers, 200 μM of each dNTP, 2.5 mM MgCl2, 1× Platinum-Taq reaction buffer, and 1.25 U PlatinumTaq (Invitrogen, Karlsruhe, Germany) were performed in a final volume of 50 μl. The reaction mixture was denatured at 95°C for 5 min, followed by 50 cycles at 95°C for 30 s, 60°C for 45 s, 72°C for 30 s, and finished with a final elongation at 72°C for 5 min.
In the next step, 45 μl PCR product was mixed for 5 min (1,200 rpm) at room temperature with 3.0 μl Streptavidin Sepharose HP (Amersham Biosciences, Freiburg, Germany) and 47 μl Binding buffer (Biotage, Uppsala, Sweden) using a thermomixer “comfort” (Eppendorf, Hamburg, Germany). Using the Vacuum Prep Tool (Biotage), single-stranded PCR products were prepared for sequencing analysis. Templates attached to the beads were washed with 70% ethanol for 5 s, denatured in 0.5 M NaOH solution for 10 s, and washed in washing buffer (Biotage) for 5 s. Then, the vacuum was switched off, and the beads were released into a PSQ 96 Plate Low (Biotage) containing 45 μl annealing buffer (Biotage) and 500 nM sequencing primer (5′-GGTTTTAAATTATGGAGTATGT-3′, nt 55039–55060 in GenBank AL161450). The samples were heated to 80°C for 2 min and then cooled down to room temperature. Pyrosequencing reactions were performed in a PSQ 96MA instrument (Biotage) according to the manufacturer’s instructions using the PyroGold SNP Reagent Kit (Biotage), which contains the enzyme, substrate mixture, and the nucleotides. Allele frequency was quantified using the SNP Software (Biotage). As described, samples were scored as heterozygous for the JAK2 mutation if the percentage of mutant alleles exceeded 5% . By definition, homozygosity was exhibited if the percentage of mutant T alleles exceeded 50% . All samples under investigation were analyzed by pyrosequencing at least in duplicate.
To analyze differences of megakaryocytic gene dosages of JAK2V617F and correlation with clinical parameters in the study group, one-way analysis of variance (ANOVA) was performed followed by Scheffé’s test for pair-wise group differences. For intra-entity analyses of JAK2V617F gene dosages and clinical parameters unpaired t tests were performed. P values≤ 0.05 were considered as statistically significant.
JAK2V617F mutation in total bone marrow cells from Ph− CMPD
A total of 258 Ph− CMPD cases were analyzed for a potential JAK2 mutation. The mutation frequency was 95% in PV (52/55), 62% in prefibrotic cIMF (58/94), 55% in advanced cIMF (39/71), and 53% in ET (20/38). Control hematopoiesis (n = 63) never displayed mutated JAK2.
Megakaryocytes in advanced cIMF showed the highest frequency of homozygous JAK2 mutation
Summary of JAK2 mutation status in megakaryocytes from Ph− CMPD isolated by laser microdissection
Heterozygous JAK2V617F mutation
Homozygous JAK2V617F mutation
Percentage of JAK2V617F homozygosity (%)
Mutated JAK2T Alleles in percent
Clinical parameters pertinent to the study group of JAK2 mutated megakaryocytes
Clinical correlates with JAK2V617F gene dosages in Ph− CMPD
Comparison of clinical parameters between the Ph− CMPD entities analyzed for megakaryocytic JAK2V617F gene dosage revealed significant higher erythrocyte counts in PV cases compared to ET (p < 0.0001), prefibrotic cIMF (p = 0.003), and advanced cIMF (p = 0.001). PV cases also showed higher hemoglobin levels and a higher hematocrit compared to prefibrotic cIMF (p = 0.006, p = 0.027, respectively) and advanced cIMF (p < 0.0001, p = 0.001, respectively) and higher hematocrit when compared to ET cases (p = 0.02). Haemoglobin levels in PVV617F and ETV617F were comparable and showed no significant difference (p = 0.079). Other clinical parameters such as leukocyte counts, platelet counts, and age showed no demonstrable differences between the Ph− CMPD under investigation.
JAK2 mutant gene dosage in megakaryocytes (MK) and the corresponding megakaryocyte-depleted bone marrow cells (MK-DBM) in 24 ET cases
Mutant JAK2 gene dosage in percent
JAK2 mutant gene dosage in MK compared to MK-DBM
∼twofold or more
Comparable or lower
JAK2 mutation status in megakaryocytes from cIMF during the course of disease
Among the eight courses under investigation, six patients presented no MF at the time of diagnosis. Three of these six showed long lasting cellular cIMF (#1–#3). A transition from heterozygous to homozygous JAK2 status occurred in patient #1, but this development excluded the megakaryocyte lineage, which remained heterozygous for the JAK2 mutation. For the other two patients (#2 and #3), heterozygosity was demonstrable in total bone marrow cells, but in patient #2, the megakaryocyte lineage showed evolution to homozygosity in 3 years of follow-up. In the remaining three patients (#5–#7), the initial cellular, prefibrotic cIMF changed to a demonstrable fibrotic stage. Patient #5 showed a homozygous JAK2 mutation already at the time of diagnosis but developed only a mild fibrosis in the following 2 years. Manifest MF and osteosclerosis rapidly developed within 21 months in patient #6 and over a time period of 14 years in patient #7, but in the latter, the JAK2 mutation never occurred so far. Patient #6 showed a decrease in the JAK2V617F gene dosage, and to note, an interferon-alpha therapy was started in the year of diagnosis.
Two of eight patients initially showed fibrosis to a different extent (#4 and #8). One of them was homozygous for the JAK2 mutation (#4) and displayed only a moderate increase in fiber deposition in a period of 4 years, whereas the other patient (#8) presented severe MF at the time of diagnosis combined with a JAK2 wild-type status.
JAK2 mutation status in total bone marrow cells and megakaryocytes during the course of disease in cIMF
Biopsy taken in (month/year)
Histopathological evaluation and diagnosis
Mutant JAK2 gene dosage in percent total bone marrow cells
Mutant JAK2 gene dosage in percent megakaryocytes
Suspicious proliferation of megakaryocytes and granulopoiesis
The histopathological evaluation of bone marrow trephines in Ph− CMPD is important for classification and supports diagnosis making particularly in manifest MF. Under this condition, bone marrow cells are regularly inaccessible for both cytological and molecular analysis (“dry tap”). Successful introduction of adequate techniques unlocked the embedded trephine for molecular analyses such as determination of JAK2 mutational status in bone marrow lineages . Up to now, solely total (unsorted) bone marrow cells from archived bone marrow samples have been analyzed for the JAK2 mutation [13–17]. Recently, megakaryocytic colony formation (megakaryocyte colony forming unit) from peripheral blood of five JAK2V617F ET patients was analyzed for the mutation, but no quantification of the gene dosage/zygosity was performed . In the current approach, we first determined the JAK2 mutation frequency by JAK2 amplification and restriction site analysis in a large series of bone marrow trephines with Ph− CMPD, which confirmed previous results [13–17]. Because atypical, pleomorphic megakaryocytes in Ph− CMPD are a predominant histopathological feature, we were interested in the mutation status of JAK2 in these cells. As described previously for analysis of megakaryocytic gene expression , 100 megakaryocytes were sufficient for DNA analysis (Figs. 1 and 2), and we used Pyrosequencing, which allows quantification of the mutated hotspot nucleotide  enabling both a discrimination between heterozygosity and homozygosity [4, 13] and designation of the mutated JAK2V617F gene dosage. Subsequently, we arbitrarily selected cases for megakaryocytic involvement of the JAK2V617F mutation with emphasis on prefibrotic and advanced cIMF and additionally in a series of PV and ET. The JAK2V617F gene dosage was significantly higher in megakaryocytes derived from fibrotic cIMF and PV than in ET (Table 1). We could demonstrate no significant difference between JAK2V617F homozygous vs heterozygous megakaryocytic lineages in ET with regard to clinical parameters. These findings are in opposite to analysis of granulocytes and unsorted bone marrow cells in ET . Recent studies correlated JAK2 mutation frequencies with clinical parameters in disease courses. Accordingly, it seems that JAK2-mutated cases in Ph− CMPD exhibit distinct differences in clinical parameters, prognosis, and survival [4, 5, 16].
Furthermore, we were interested in megakaryocytes as the potential predominant cellular lineage affected by the JAK2 mutation in ET, as a previous study identified two patients with JAK2V617F in platelet-derived cDNA but wild-type JAK2 in the corresponding mononuclear cells . Although about 80% of ET cases under investigation displayed comparable or even lower levels of mutated JAK2 gene dosages in megakaryocytes than the corresponding megakaryocyte-depleted cellularity, actually 5/24 ET cases revealed a striking higher JAK2V617F gene dosage in megakaryocytes. On top, out of this heterogeneous pattern, one ET case obviously acquired the JAK2 mutation solely in the megakaryocytic lineage because the surrounding megakaryocyte-depleted bone marrow cells showed the wild-type of JAK2. Referring to this, a recently published study demonstrated that in PV, the JAK2V617F mutation could have been acquired exclusively in a clone of the erythroid linage . The questions remains to be resolved if also cIMF has a “particular” JAK2V617F lineage.
Megakaryocytes are supposed to play a crucial role in MF development, and in our study group, megakaryocytes displayed higher frequencies for homozygous JAK2V617F in fibrotic cIMF (60%) than prefibrotic cIMF (33%, Table 1), although the difference was not statistically significant. We  and others [23–25] have previously demonstrated that the JAK2 mutation is not required for MF development, which was confirmed by the identification of two representative cIMF disease courses with JAK2 wild-type status but severe MF (#7 and #8, Table 4). Therefore, we did not analyze wild-type vs JAK2V617F but dynamics of different megakaryocytic JAK2V617F gene dosages in six representative cIMF disease courses with known JAK2 mutation (Table 4). Putting aside the fact that there was no influence of the gene dosage on the stages of bone marrow MF, we observed some interesting findings in the mutation status of megakaryocytes when compared to their corresponding total bone marrow cells. It has been demonstrated that in some cases, transition to homozygosity in Ph− CMPD is a matter of time since cells homozygous for the mutation will, little by little, preponderate the malignant clone [4, 15]. In our study, we identified a patient who showed a doubling of the mutated gene dosage and transition to JAK2V617F homozygosity in total bone marrow cells but surprisingly not in megakaryocytes (#1, Table 4). Thus, we conclude that the evolution from heterozygous to homozygous JAK2V617F did not affect all myeloid lineages. This might reflect the complexity of different lineages involved in the mutation and the fraction of cells already having undergone mitotic recombination. However, it cannot be excluded that homozygosity may develop later on or that a minor fraction of megakaryocytes have already undergone transition to homozygosity. Nevertheless, even in case of a small homozygous subfraction, they are obviously of minor relevance when compared to the changes of the total bone marrow cells.
The ∼20% decrease in the JAK2V617F gene dosage in cIMF patient #6 was most likely an equivalent to a therapy-related effect on the mutated clone including the megakaryocytic lineage (Table 4). Rapid development of severe MF might be secondary to apoptotic cells, especially megakaryocytes. Indeed, there are suggestions to use JAK2V617F quantitative detection assays as a therapy-monitoring tool , but further data need to be generated for a sufficient conclusion.
We conclude that (1) megakaryocytes might be the predominant or even the exclusive cellular lineage that acquires the JAK2V617F mutation in ET, (2) JAK2V617F ET-derived megakaryocytes were associated with a PV-like phenotype, and (3) JAK2V617F evolution to higher mutant gene dosages and homozygosity represent a dynamic and complex process in megakaryocytes and other myeloid lineages independent of the development of bone marrow fibrosis.
The authors gratefully acknowledge the technical assistance of Ms. Sabine Schroeter. The authors would like to thank Dr. Masyar Monazahian, Niedersächsisches Landesgesundheitsamt, Hannover, Germany for the opportunity to perform the pyrosequencing assay on the PSQ 96MA instrument. Research grants: Deutsche Krebshilfe, Dr. Mildred Scheel Stiftung 10-2191 (OB, HK), Deutsche Forschungsgemeinschaft-DFG BO 1954/1-1 (OB, HK).