Targeted Cancer Exome Sequencing Discovers Novel Recurrent Mutations In MPN

The discovery of the JAK2V617F mutation in 2005 [Kralovics R, N Engl J Med 2005] represented a major breakthrough in the understanding of the molecular pathogenesis of Philadelphia chromosome negative chronic myeloproliferative neoplasms (MPN). Nevertheless several observations suggest that the JAK2V617F mutation may not be the disease funding mutation, at least in most instances. Therefore, a great deal of effort is ongoing with the aim to identifying novel genetic lesions contributing to the disease pathogenesis. The two major theoretical and technical drawbacks to the identification of new somatic mutations are represented, respectively, by the huge number of genes potentially involved in tumorigenesis of MPN and by the availability of a “pure” germline control DNA. Buccal swabs and saliva have been generally considered as readily available sources of DNA of non-hematopoietic origin, but detection of the JAK2V617F mutation in at least some of these samples indeed suggested the presence of myeloid cell contamination [Levine RL, Cancer Cell 2005]. So, in order to discover novel mutations in MPN using upfront technologies based on next-generation sequencing (NGS) we designed a “cancer exome” capture panel of 2000 unique genes and microRNAs. This panel was used to capture libraries generated from genomic DNA extracted from granulocytes and in vitro expanded CD3+ T-lymphocytes as germline control, in a cohort of 20 MPN patients. These captured libraries were than massively sequenced using the Roche 454 FLX platform. DNA samples had been collected at the diagnosis of PV in 9 subjects and PMF in 6 subjects, while the remaining 5 DNA samples were from 5 of the 9 PV patients at the time they evolved to post-PV myelofibrosis.

After extensive bioinformatics analysis and multiple control adjustments, we finally produced a list of 171 novel “true” somatic mutations occurring in genes and microRNAs coding regions of those MPN samples; some of these mutations have been already described in MPN, whereas novel variants represent the vast majority. Despite patients harbored different numbers of somatic mutations, spanning from four to twenty-one variants, only 22 genes appear recurrently mutated. It is worth of notice the acquisition of additional mutations and/or the occurrence of loss of some mutations at the time of disease evolution from PV to a post-PV Myelofibrosis in the five patients for whom samples were available at both disease phases. Some of them, either acquired (NTRK1, PRDM2, BRCA2 and BARD1) or lost (APC, CARS, MLL3 and FAT2) had been found also in another PV or PMF sample.

To test the recurrence of these mutations, we screened a different cohort of 189 patients composed by PMF (91 samples), PV (50 patients) and post-PV Myelofibrosis (48 samples) by Ion AmpliSeq technology on an Ion Torrent PGM platform. Deep amplicon sequencing of granulocytes DNA achieved a sample median of 1000-fold coverage. Excluding JAK2, MPL, IDH2, ASXL1 known variants, for 7 genes (SCRIB, MIR662, BARD1, TCF12, FAT4, DAP3, NRAS) we demonstrated in MPN a global mutation frequency greater than the 3%. Whereas some new variants need functional validation to prove causal mechanisms, some other mutations have a well-known pathogenic role in solid cancers but here are described for the first time in MPN.