P323: uncovering subclonal metabolic vulnerabilities in acute lymphoblastic leukemia at single cell resolution

Background: Despite improvements of therapy regimens for acute lymphoblastic leukemia (ALL), a significant number of patients experience relapse or do not respond to treatment, highlighting the need for continued research into the underlying mechanisms and the development of new treatments. It has been suggested that metabolic alterations are key drivers of therapy resistance. While most of the current understanding of ALL is based upon assays that investigate the characteristics of bulk samples, it is important to characterize the disease at the single cell level to deconvolute tumor heterogeneity that may drive relapse. Aims: We aim to identify metabolic features that are unique to different subtypes of ALL. Furthermore, we will investigate mechanisms of metabolic resistance of ALL using a novel assay that integrates transcriptomics, metabolism, and clonal hierarchy at the single cell level. Methods: Bone-marrow aspirates from ALL patients were stained with up to 19 surface markers and metabolic dyes for intracellular pH, mitochondrial potential, autophagy and reactive oxygen species, and were analyzed by spectral flow cytometry. A selected BCR-ABL positive ALL case sampled at nine timepoints throughout the therapy, was index-sorted in 384-well plates. Libraries were processed according to SmartSeq2 protocol with additional amplification of clinically relevant driver mutations. Index-sorting enables to connect the information about metabolic fluorescence with the transcriptome of each individual cell. Results: Bone marrow aspirates of B-ALL patients covering multiple subtypes were subjected to high parametric spectral flow cytometry analysis. Leukemic cells could be separated from healthy counterparts by metabolic features. Furthermore, differences in metabolism of leukemic cells could be identified between subtypes. For example, while MLL-rearranged samples had a higher intracellular pH, BCR-ABL positive samples displayed higher activity in autophagy. From a BCR-ABL+ ALL case sampled at nine timepoints throughout the therapy we subjected 3840 to our single cell multiomics platform, of which 3456 cells were used for downstream analysis. Leukemic cells, identified by the presence of the BCR-ABL gene fusion, were traceable at most of the subjected timepoints. Based on mitochondrial variants, multiple subclones could be identified within the leukemic population. During therapy a subclone with a T315I mutation evolved leading to TKI-therapy resistance. Ancestors of this clone showed changes in glucocorticoid metabolism and transcriptional state already at initial diagnosis. Summary/Conclusion: In a cohort of primary ALL samples we were able to trace differentiation and metabolism of leukemic cells with high parametric flow cytometry. Genetic subtypes were associated with distinct metabolic changes in intracellular pH and autophagy. We determined clonality, metabolism and transcriptome in single cells of a patient with BCR-ABL positive ALL over multiple timepoints. Subclonal hierarchies were established based on the BCR-ABL gene fusion and mitochondrial mutations. While the major subclone responded to therapy, a minor subclone emerged with a relapse driving T315I mutation. Precursors of this subclone had already an altered metabolic state at initial diagnosis. Keywords: BCR-ABL, Acute lymphoblastic leukemia, Flow cytometry, Clonality