The CML and the ALL inducing isoforms of BCR::ABL1 differ for the induction of genomic instablity

Abstract

Abstract There are two major isoforms of BCR::ABL1, the Philadelphia chromosome-positive acute lymphoblastic leukemia (Ph+ ALL) specific p185BCR::ABL1, and the p210BCR::ABL1, which defines 95% of chronic myeloid leukemia (CML). These two isoforms differ in the presence of the oncogenic pleckstrin- and DBL-homology domains (DH and PH) in p210BCR::ABL1. The p210BCR::ABL1 induces genomic instability (GI) by double-strand breaks (DSB). The DNA damage and repair response (DDR) repairs DSBs. BCR::ABL1 compromises the fidelity of DDR by promoting the repair of DSB through single-strand annealing (SSA), a mutagenic pathway, and through non-homologous end joining (NHEJ). Both DDR mechanisms may not be sufficiently suppressed by tyrosine kinase inhibitors (TKI), resulting in imprecise repair and mutations. As a model, we used early passages of thirteen long-term cultures derived from patients (PD-LTCs) with acute lymphoblastic leukemia (ALL). Five PD-LTCs were from Ph+ ALLs with the associated p185BCR::ABL1, one of them with the T315I mutation. The two PD-LTCs from lymphoblastic CML blast crisis (CML l-BC) had the p210BCR::ABL1 fusion protein. The Ph-like PD-LTC exhibited an ETV6::ABL1 fusion protein. Two PD-LTCs harbored the t(1;19), which encoded the E2A::PBX1 fusion protein. The t(11;19) translocation encoding MLL::MLLT1 characterized one PD-LTC. We used the two PD-LTCs with no translocation and a complex karyotype as a reference. These cells represent the closest model to primary ALL cells as they are genomically and immunologically stable and did not undergo the classical crisis of a transformed cell line. They also enabled the determination of a relationship between GI and the response to TKI therapy in Ph+ ALL, as each of the Ph+ and Ph-like PD-LTCs has a different response to TKI. For the genome-wide measurement of the DSBs, we focused on INDUCE-seq, a technique we have recently developed. It allows for defining every single DSB at a molecular level, followed by bioinformatic analysis of DSBs. In this way, it is possible to define the absolute number of DSBs, their recurrency, and the number of DSBs per chromosome and their role in response to therapy and leukemogenesis. We aimed to determine the role of GI induced by rearranged ABL1 in the pathogenesis of Ph+ and Ph-like lymphoblastic leukemia and the response to TKI. We also wanted to compare the two major BCR::ABL1 isoforms, p185BCR::ABL1 and p210BCR::ABL1, regarding the induction of GI. We found that PD-LTCs from Ph+ ALL with their p185BCR::ABL1 induced a higher rate of GI by DSB than those from CML l-BC with their p210BCR::ABL1. We observed a higher absolute number of DSBs normalized to the amount of DNA extracted in p185BCR::ABL1-positive PD-LTCs (mean of ~89K) than in p210BCR::ABL1-positive PD-LTCs (mean of ~47K) and every other PD-LTC (mean of ~42K). ETV6::ABL1 induced a total number of DSB close to p185BCR::ABL1 (mean of ~ 75K) in the Ph-like PD-LTC. For both BCR::ABL1 isoforms and the Ph-like PD-LTC, the recurrency (the rate of DSBs in single genes) was higher than in other PD-LTCs, with a higher rate in p185BCR::ABL1-positive PD-LTCs than in those harboring p210BCR::ABL1. There was a relationship between sensitivity to TKI and the recurrency of DSB in the TKI-sensitive PD-LTCs. The lower the recurrency, the higher the sensitivity to TKI. The p185BCR::ABL1 fusion strongly induced DSBs in chromosomes 5 and 8. As no the induction of DSBs in chromosomes 5 and 8 was more likely related to the leukemogenesis of Ph+ ALL. Our still preliminary data demonstrate the significance of GI for the leukemogenesis and the response to TKI of lymphoblastic leukemia with the rearrangement of ABL1. It is not yet clear whether GI may be independent of ABL1 kinase activity, a target of TKI. It has to be confirmed by clinical studies.