The JAK2 V617F mutation is detectable in granulocyte populations at greater than two copies per cell among individuals with myeloproliferative disorders

A recent letter in Leukemia¹ reported the detection of the JAK2 V617F mutation at very low levels in peripheral blood of healthy donors. This finding is consistent with our own data generated using TaqMan (fluorescence-based real-time PCR) and minor groove binding probes designed for the quantitative determination of DNA copy number of JAK2 V617F.²

Apart from not using a plasmid standard curve containing the mutant allele to determine absolute copy number of JAK2 V617F, the characteristics of the previously reported assay appear very similar to our own.² We found that our assay, which does not employ locked nucleotide technology, exhibited good reproducibility, efficiency and the correlation coefficient of the standard curve averaged 0.97. The assay showed 100% concordance with sequence-based typing (SBT) for detecting the presence of the mutation in samples with greater than 1000 copies (n=14), and detected low levels of the mutation in three samples that showed no mutation by SBT. We also detected the presence of the mutation to less then 0.05% clinical dilution (30 copies), with a linear range extending from 10 and 300 000 copies per reaction (Figure 1). By combining this assay with another TaqMan assay for determining cell copy number,³ we have been able to determine the average number of JAK2 V617F mutant copies per cell and per 20 ng of DNA in absolute terms from granulocyte fractions (Figure 2). Quantitative determination of JAK2 V617F by either means was extremely low among healthy controls (n=18), and we assumed this to represent non-specific amplification of the wild-type allele, as shown in Figure 1. Among patients presenting with suspected myeloproliferative disorders (n=131) some values were extremely high and included six measures greater than an average of 2.5 copies per cell from individuals. Although the veracity of this finding should be confirmed by an alternative approach, these data suggest that disease evolution may involve further gene duplication associated with genetic instability. Our data contribute to on-going questions regarding the interpretation of variations in JAK2 V617F copy number at the cell population level. Levels of mutation that are significantly above those seen in healthy controls are clearly of clinical relevance, but beyond providing an accurate measure of clonal proliferation, the potential contribution of this measure to inform clinical management remains to be demonstrated.

Figure 1

Clinical dilutions of DNA from a mutation positive individual with DNA from a healthy control. The assay demonstrated sensitivity to 0.05% mutation. Amplification in healthy control samples (n=18) did not occur before 36 cycles or 30 copies. Determining the specificity of the assay is hampered by fact that we do not know whether healthy controls also carry the mutation. However, we can say that amplification of the mutation in the healthy controls occurred only after the number of cycles required to detect a 0.05% dilution of the positive sample.

Figure 2

Twenty nanograms of input DNA (n=131 samples), as quantified by NanoDrop spectrophotometry, was used in two real-time PCR assays to quantitatively determine JAK2 V617F levels and cell numbers (based on two copies of the HGH gene being amplified per cell), respectively (top left panel). An average of 9887 cells (s.d.=3291; CV=33%) per 20 ng of input DNA was measured for all samples, in comparison to the 3333 cells which would be expected to be measured in a perfect DNA extraction and amplification system (assuming one diploid cell carries 6 pg of DNA). Greater cell numbers were measured in samples from possible MPD cases (mean=10439 cells) compared with controls (mean=5840 cells). JAK2 V617F levels expressed by either method produced a similar distribution of the data (bottom panels). However, the read-out from the internal locus control assay is more biologically meaningful (i.e., copies/cell) than an absolute number of copies, and the clinical significance of changes in this measurement with treatment may be more easily appreciated. The top three measures of JAK2 V617F levels were from male individuals with grossly hypercellular bone marrow, a diagnosis of polycythemia vera and who were all treated with hydroxy urea.