Detection of chemotherapy-resistant patient-derived acute lymphoblastic leukemia clones in murine xenografts using cellular barcodes

Abstract Clonal heterogeneity fuels leukemia evolution, therapeutic resistance and relapse. Upfront detection of therapy-resistant leukemia clones at diagnosis may allow adaptation of treatment and prevention of relapse, but this is hampered by a paucity of methods to identify and trace single leukemia-propagating cells and their clonal offspring. Here, we tested methods of cellular barcoding analysis, to trace the in vivo competitive dynamics of hundreds of patient-derived leukemia clones upon chemotherapy-mediated selective pressure. We transplanted Nod/Scid/Il2Rγ−/– (NSG) mice with barcoded patient-derived or SupB15 acute lymphoblastic leukemia (ALL) cells, and assessed clonal responses to dexamethasone, methotrexate and vincristine, longitudinally and across nine anatomical locations. We demonstrate that chemotherapy reduced clonal diversity in a drug-dependent manner. At end-stage disease, methotrexate-treated patient-derived xenografts had significantly fewer clones compared to placebo-treated mice (100±10 versus 160±15 clones, p=0.0005), while clonal complexity in vincristine- and dexamethasone-treated xenografts was unaffected (115±33 and 150±7 clones; p=NS). Using tools developed to assess differential gene expression, we determined whether these clonal patterns resulted from random clonal drift or selection. We identified five clones that were reproducibly enriched in methotrexate-treated patient-derived xenografts, suggestive of pre-existent resistance. Finally, we found that chemotherapy-mediated selection resulted in a more asymmetric distribution of leukemia clones across anatomic sites. Altogether, we demonstrate that cellular barcoding is a powerful method to trace the clonal dynamics of human patient-derived leukemia cells in response to chemotherapy. In future, integration of cellular barcoding with single-cell sequencing technology may allow in-depth characterization of therapy-resistant leukemia clones and identify novel targets to prevent relapse.