Mosaic Distribution of RAS Pathway Mutated and Non Mutated Cells in Bone Marrow of Juvenile Myelomonocytic Leukemia.

Abstract 2818

Juvenile myelomonocytic leukemia (JMML) is a rare myeloproliferative disorder of early childhood characterized by excessive proliferation of granulocytic and myelomonocytic cells. So far, about 85–90% of patients harbor mutations in genes (RAS, PTPN11, CBL, NF1) involved in the RAS signaling pathway and 25% of cases show monosomy of chromosome 7. In 10–15% of patients with JMML no recurrent mutations have been identified. In JMML, evidence of mutated cells in the myeloid lineage is well known; however, to which extend cell populations of other hematopoietic lineages harbor the same mutation is not well understood. We calculated the mutated allele frequencies of PTPN11, KRAS and CBL mutations of 10 patients with JMML using amplicon ultra deep sequencing 454 Roche Technology. KRAS and PTPN11 mutated allele frequencies in the total bone marrow (BM) varied in a range from 37.40% to 51.58%, and no statistical difference was identified in the variants allele read counts of the latter two genes. The only patient analyzed with a CBL mutation showed a mutated allele frequencies of nearly 100% pointing to homozygosity of the mutation and suggesting loss of the non-mutated CBL allele through somatic uniparental disomy of chromosomal region 11q23. Assuming occurrence of PTPN11 and KRAS mutations in heterozygosity (presence once per diploid genome), a mutated allele frequency of approximately 50% would imply that all cells were mutated (100% of cells). Along this line, our analysis of mutation frequencies, largely below 50%, pointed to a mosaicism of mutated and non-mutated cells in the BM of JMML patients. To further analyze the spectrum of mutations in BM of JMML, we sorted different hematopoietic subpopulations from 6 patients carrying a PTPN11 mutation. We choose different combinations of surface markers: CD34 and CD38 for stem and progenitor cells, CD33, CD14, CD15 and CD16, CD11b for myeloid lineages at different stages of maturation, CD3 and CD19 for T and B cells, respectively. Analysis of sorted cells revealed high mutated allele frequencies for all maturation stages of the myeloid lineage with more than 90% heterozygous mutated cells. Presence of high mutated allele frequencies were also identified in hematopoietic stem and progenitors cells and in B-cell lineage. Interestingly, in the T-cell lineage, a low mutated allele frequency was detected ranging from 9.4% to 29,35%, this finding suggesting that about 20% to 60% of T cells in the BM harbor PTPN11 mutations. Furthermore, with the aim of detecting still unknown mutations, associated with JMML we performed whole exome sequencing of BM cells of 4 JMML patients lacking any of the currently known mutations. In conclusion all hematopoietic lineage cells carry PTPN11 mutations in different percentages, indicating a mosaic distribution of mutated and non-mutated cells in the whole BM of JMML patients. We studied the power function of a binomial test to understand the probability of detecting a real deviation of variant allele read count from the hypothetical 50%, as expected for 100% of mutated heterozygous alleles. Moreover we provided a sensitivity analysis of the possible p-values when the 50% hypothetical model is true. Also using this model, a scenario of mosaicism appeared in the various subpopulations of JMML BM. The presence of non-mutated apparently healthy cells in the BM of JMML patients may hold a promise for exploring therapies that could interfere only with the mutated population sparing healthy cells in the myeloid progenitor cells populations.