Collaborative Pathways that Lead to Leukemia

Division Of Basic Sciences - Nci

Investigators

Linked publications & trials

Abstract

Using NP23 mice,we identified frequent acquired mutations in progenitor B1 cell ALL in the Bcor and Jak1/2 genes. Subsequently, we used CRISPR to introduce Bcor mutations in primary WT and NP23 BM cells; these cells were transplanted into recipient mice and verified collaboration between NP23 and Bcor in vivo. The leukemias also acquired Jak pathway mutations, suggesting that NP23, Bcor, and Jak collaborated to produce pro-B1 ALL. These results were published in FY2020 (PMID: 30992267). We previously demonstrated that spontaneous mutations of IDH2 occur in NHD13 leukemias. These mutations occur at R140Q; homologous residues are mutated in human leukemia. We crossed IDH2 R140Q transgenic mice with NHD13 mice; the offspring develop a form of early T cell precursor (ETP) leukemia that resembles the human disease in terms of clinical presentation, immunophenotype, gene expression profile, and collaborative mutations. In terms of molecular mechanism, the IDH2R140Q mutant mice produce the oncometabolite 2HG; consistent with overproduction of 2HG, the leukemic cells show aberrant methylation of genes required for normal thymocyte development. Finally, a specific inhibitor of mutant IDH2 (AG-221) inhibits the growth of these ETP cells in vitro. A manuscript describing these findings was published in FY2021 (PMID: 34321240). Both NP23 and NHD13 mice develop a wide variety of leukemia at 9-14 months of age. Surprisingly, 100% of the NP23-NHD13 double transgenic mice developed acute myeloid leukemia (AML) within 3 months. The leukemias were characterized by extraordinarily high WBC and replacement of the thymus with AML cells; the percent of malignant myeloid cells in the thymus was often higher than the bone marrow (BM). These findings led to the intriguing hypothesis that the AML in NP23-NHD13 mice arose in the thymus, as opposed to the BM. To investigate this possibility, we transplanted CD4-/CD8- double negative (DN) thymocytes from the thymus of a NP23-NHD13 mouse. All mice developed AML within 26 days, indicating that the AML was aggressive and transplantable, and could be transmitted by DN thymocytes. Fractionating DN thymocytes into DN1-DN4 sub-populations revealed that AML initiating cells were found in the DN1 and DN2 compartments. Taken together, these results demonstrate that NP23-NHD13 thymic progenitors retain myeloid and erythroid potential and are potently leukemogenic, leading to the intriguing hypothesis that some human AML might originate in the thymus. A manuscript describing these findings was published in FY2020 (PMID: 31748606). Ongoing experiments are focused on assessing the leukemic potential of combining Bahcc1 mice (see project 1) with NHD13 transgenic mice. Mini-chromosome maintenance component 2 (Mcm2) is a DNA replication licensing factor that is part of the Mcm2-7 complex which functions as a DNA helicase, unwinding genomic DNA at the replication fork. Not surprisingly, homozygous deletion of Mcm2 is lethal. However, insertion of a cre cassette into the 3' UTR of Mcm2 leads to 50% reduction in Mcm2 protein, and cells with two copies of the cre knock-in allele express only 20-30% as much Mcm2 protein compared to wild-type cells. Despite the diminished Mcm2 protein levels, mice with two copies of the Mcm2cre allele are born at normal Mendellian ratios, are not growth-retarded, and are indistinguishable from wild-type littermates at two months of age. Beginning at 2-3 months of age, the mice become ill, and invariably die from pre-T lymphoblastic leukemia/lymphoma (pre-T LBL). Copy number alteration (CNA) analysis reveals a pattern of gains and losses, predominantly losses 10-1000 kb in length. Notably, there is a recurrent constellation of losses, including biallelic deletions of Pten, Tcf3 (E2a) and Dnmt3a, and mono-allelic deletions of the amino-terminus of Notch1. This constellation of cooperative deletions fits a model (supported by published experiments with Pten, Tcf3, Dnmt3a) in which Dnmt3a deletion leads to increased stem cell self-renewal, Tcf3 deletion blocks thymocyte differentiation, Pten deletion leads to hyperproliferation, and deletion of the amino terminus of Notch1 leads to ligand independent growth. All of these genes except TCF3 are frequently mutated in human T-ALL; NOTCH1 being the single gene most commonly mutated in human T-ALL. Although TCF3 is not frequently deleted in human T-ALL, TCF3 is functionally inactivated by inappropriate expression of TAL1/SCL and LMO1/2 proteins (EMBO J 16:2408-19; Nature Immunol 1:138-44) in 25-50% of human T-ALL patients (Cancer Cell 1:75-87), underscoring the relevance of TCF3 inactivation in human T-ALL. Mice that express a NUP98-HOXD13 (NHD13) transgene develop myeloid, T-cell, and B-cell leukemia. Crossing the NHD13 transgene onto an Mcm2cre/cre background led to B-cell precursor (BCP) ALL in a subset of Mcm2cre/creNHD13+ mice. CNA analysis of these BCP-ALL revealed consistent deletions in Pax5, gains of a region bounded by Nup214 and Abl1, and bi-allelic loss of Ptpn1. The gains of Nup214 and Abl1 led to generation of a Nup214-Abl1 fusion gene, similar to that seen in some human T-ALL and BCP-ALL patients. PTPN1 deletions have not been reported in human BCP-ALL, however, deletions of the closely related PTPN2 co-occur with NUP214-ABL1 fusions, and PTPN2 was identified as a negative regulator of the NUP214-ABL1 kinase (Nat Genet 42:530-5, 2010). This constellation of cooperative losses and gains fits a model in which the NHD13 transgene leads to increased stem cell self-renewal, the Pax5 deletion leads to a block in B cell differentiation, the Nup214-Abl1 fusion leads to hyperproliferation, and the Ptpn1 deletion enforces hyperproliferation. Similar to the findings for T-ALL, these genes and pathways have been highlighted as being important for human BCP-ALL (see review by Mullighan and Hunger, Blood 125:3977-87). Sequencing of 323 independent junctions from 91 tumors revealed that mononucleotide repeats were enriched near the breakpoint junctions, consistent with recent observations that increased replicative stress is associated with DNA DSB at sites of mononucleotide repeats (Cell 174:1127-42, 2018). Overall, this Mcm2 deficiency leads to a unique mutator phenotype, characterized by copy number gains/losses of 50-1000 kb. A manuscript describing these findings was published in FY2020 (PMID: 31622281). In theory, this mutator phenotype could be used to identify constellations of mutations in other forms of cancer if they lived for 3 months. if Mcm2cre/cre mice could be protected from the highly penetrant pre-T LBL, we might uncover additional malignancies triggered by the Mcm2 deficiency. Therefore, we crossed Mcm2cre/cre mice onto a nu/nu background and demonstrated that these mice are indeed protected from development of pre-T LBL, living a median of 8 months as opposed to 3 months. However, the vast majority of Mcm2cre/cre:nu/nu mice develop B cell ALL, beginning at approximately 9 months of age. We are now crossing the Mcm2cre/cre mice onto scid/scid and RAG2 KO backgrounds to determine if mice that lack VDJ recombination will be "protected" from both T and B cell malignancies.