Identification of New Leukemia Associated Gene Mutations by Whole Genome Sequencing of DNA From Two Siblings with Congenital Neutropenia and Secondary AML

Abstract 12

To identify the pattern of genetic aberrations, which may promote leukemia development in patients with severe congenital neutropenia (CN), we have performed a whole genome sequencing (WGS) of DNA samples from myeloid leukemic cells of two affected siblings suffering from CN. Both children harbored ELANE gene mutations. The father of the children demonstrates somatic mosaicism for the ELANE mutation and has no severe neutropenia. For WGS we used Complete Genomics technology (Complete Genomics. Inc, Mountain View, CA.). More than 90 % of genomes were sequenced at high quality with minimum coverage of at least 20-fold. As an example, 3.355.399 single nucleotide variants (SNVs) were identified in DNA isolated from leukemia blasts of one CN patient. The following filters were used to identify mutations in the leukemic cells from the two patients: 1) Non-synonymous SNVs in coding sequences only (9288 SNVs), 2) 54 healthy individuals sequenced by Complete Genomics (557 SNVs), 4) five members of one family from the same ethnic area (healthy parents, one cyclic neutropenia patient and her healthy brother, 471 SNVs), 5) five family members of the affected two children: parents and three healthy siblings (two healthy sisters and one heatlthy brother). Remaining SNVs presented in the two affected children were subsequently analysed using in silico prediction software Polyphen 2, which predicts possible impact of an amino acid substitution on the structure and function of human proteins. Fourteen SNVs with predicted damaging effects on the protein function were used for further analysis. All candidate SNVs were validated by Sanger sequencing. We detected nine inherited candidate SNVs presented in the two affected children but not in healthy siblings. The SNV in the ELANE gene (c.452G>A p.C151Y, dbsnp.129:rs57246956) was inherited from the father. Novel SNVs inherited from the father were as follows: in the TCTE1, FAM135A, M6PR, C20orf144 and PTPN23 genes. Only three SNVs were inherited from the healthy mother (in BLOC1S1, DUS3L and KIAA1543 genes). All SNVs were heterozygous. We also found 5 sporadic SNVs presented in leukemia sample of one CN patient only, but absent in his DNA sample from an earlier time point of CN diagnosis. These are heterozygous SNVs in the CSF3R, ACAP2, GRM1, LASS3, and RUNX1 gene. All five gene mutations might be involved in leukemogenesis. Interestingly, both affected patients had somatic mutation in the RUNX1 gene at the same nucleotide position (c.415C>G, p.R139G in sick brother and c.415C>T, p.R139* in sick sister). In summary, we identified candidate genes that may be relevant for leukemogenesis in CN patients. Our study also establishes WGS as an unbiased method for discovering leukemia-initiating mutations in previously unidentified genes that may respond to targeted therapies