B-, T-, and NK-cell lineage involvement in JAK2V617F-positive patients with idiopathic myelofibrosis

Abstract

An acquired JAK2^V617F mutation has been found in myeloid cells from most patients with chronic idiopathic myelofibrosis (IM), but whether it occurs in a common myelo-lymphoid, rather than a myeloid-restricted, progenitor cell is still debated. Using a sensitive ARMS assay for the quantitative assessment of JAK2^V617F cDNA, we detected the mutation in purified B, T and NK cells from about half of 12 patients studied. These results indicate that involvement of the lymphoid lineage in IM may be more frequent than previously supposed.

Patients with chronic myeloproliferative disorders (MPD) harbor a recurrent JAK2 mutation1 that has been consistently found in their granulocytes but not in control tissues. In cases in which T lymphocytes were analyzed, they resulted uniformly wild-type;2,3 likewise, the mutation could not be found in CD19 cells purified from three homozygous cases of MPD using a direct sequencing approach.4 These data would suggest that the cell target for the JAK2^V617Fmutation is a myeloid-restricted progenitor. However, this idea is challenged by the finding that cells with the hematopoietic stem cell (HSC) phenotype (CD34, CD38, CD90, Lin) from patients with polycythemia vera (PV) actually harbored the mutation. 5 Furthermore, Ishii et al.6 recently reported that B and T cells from, respectively, 2/8 and 1/8 PV patients were JAK2^V617F mutated. Interestingly, the mutation was also found in B, T, and NK cells of one patient with familial MPD.7

We analyzed 12 patients with IM, at variable times from diagnosis, searching for the presence of the JAK2^V617F mutation in their lymphoid cells. All patients gave informed consent to these investigations. Peripheral blood (PB) granulocytes were collected by density gradient centrifugation to ≥95% purity by visual inspection of cytosmears; CD3 and CD19 lymphocytes were obtained by direct immunomagnetic selection to a purity ≥98% by FACS re-analysis. In five patients, the purity of CD56/CD3 NK cells and CD3/CD56 T-lymphocytes obtained using an immunomagnetic depletion procedure was ≥93% and ≥95%, respectively. PB CD34 cells were purified to ≥97% by direct immunomagnetic selection (Miltenyi Biotec, Germany). The percentage of JAK2^V617F cDNA in these cell fractions was measured using an ARMS procedure.8 Briefly, RNA was reverse transcribed and amplified using fluorochrome-labeled mutation-specific primers; amplicons were resolved by capillary electrophoresis, and the ratio of JAK2^V617F to JAK2 (JAK2^V617F+JAK2) cDNA was calculated.8

We observed a significant, although heterogeneous, pattern of lymphoid cell involvement in IM patients (Table 1). In particular, the JAK2^V617F mutation was detected in the CD19 cells of 7/12 patients (58%), in the CD3 cells of 5/12 (42%), and in the CD56 NK cells of all the five cases in which they were evaluated. Remarkably, the JAK2^V617F ratio measured in these highly-purified lymphoid cell preparations was high enough, when compared to the corresponding value in granulocytes, to exclude the possibility that the presence of the mutation may have been due to a few contaminating granulocytes. The presence of the JAK2^V617F mutation in lymphoid cells was not obviously associated with the highest mutational load in myeloid cells; for example, patient #1 had a JAK2^V617F ratio of 100% in granulocytes, while both CD9 and CD3 were wild-type, and conversely patient #8, in whom a JAK2^V617F ratio of 26% was found in granulocytes, had a ratio of 30% in CD19 cells and wild-type CD3. Patient #11, who had a 6-year long history of the disease, was particularly interesting and presented a 100% ratio in all the different cell fractions examined.

Table 1.The JAK2^V617F mutant allele ratio in different hematopoietic cell fractions purified from the peripheral blood of patients with idiopathis myelofibrosis.

We also found that CD34 cells harbored the JAK2^V617F mutation at ratio values roughly similar to those in granulocytes. Although immunomagnetically selected CD34 cells contain both CD38 and CD38 cells, and thus do not truly reflect a HSC phenotype, nonetheless this observation suggests that the target cell in IM (and PV)6 may be a common myelo-lymphoid progenitor,9 and indirectly makes it unlikely that the phenotypic pleiotropy of JAK2^V617F mutated MPD patients results from different cellular targets.

Overall, the current data are in line with those recently reported by Delhommeau et al.9 in ten IM patients. These authors found the JAK2^V617F mutation in B-cells of 60% of the patients, in 25% in case of T-cells, and 63% of NK cells. The variable frequency of lymphoid cells involved in the JAK2^V617F mutant clone might be also explained by the long lifespan of these cells, especially T-lymphocytes, most of which actually pre-existed the event(s) leading to the acquisition of the JAK2^V617F mutation. Consistently, JAK2^V617F mutant T cells were found in all IM (and PV) patients evaluated using the fetal thymus organ culture (FTOC) assay.9

Larger studies are clearly needed to obtain a consistent figure of the incidence of JAK2^V617F mutation in lymphoid cells, but from these data it appears to be not sporadic; specifically, the low-sensitivity direct sequencing approach employed in the first studies might have been the reason for the under-estimation of this phenomenon. 2–4 The consequences, if any, of the involvement of B cells, T cells and NK cells by the JAK2^V617F mutation are totally unclear so far, but one is tempted to speculate that it might underlie some of the immunologic abnormalities manifested by a significant proportion of IM patients.10