Alternating cycles of quiescent and proliferative cell states determine stemness and leukemia-initiation capacity in acute lymphoblastic leukemia

Abstract Background and significance. Stemness in acute myeloid leukemia (AML) is determined by a clonal hierarchy with rare, phenotypically defined leukemia-initiating cell (LIC) at its apex (Lapidot 1994, Bonnet 1997). Given that LICs in AML are drug-resistant and initiate relapse of refractory disease, the identification of an LIC population in B-ALL would be consequential. However, unlike AML, B-ALL does not evolve from a clonal hierarchy and LIC populations have remained elusive in B-ALL (Kelly 2007, Le Viseur 2008, Rehe 2013). Results. Time-lapse studies of single patient-derived B-ALL cells revealed that most B-ALL cells were continuously proliferating (steady), while subpopulations underwent periodic transitions between quiescent and proliferative cell states (alternating). Gene expression studies of 'steady’ and 'alternating’ cells revealed a dominant and reciprocally repressive transcriptional program driven by MYC in 'steady’ and BCL6 in 'alternating’ B-ALL cells. To dissect how MYC and BCL6 dynamics govern transitions between proliferative and quiescent cell states in B-ALL, we engineered patient-derived B-ALL (PDX) by CRISPR and HDRT with dual-reporter knockin alleles expressing C-terminal MYC–mNeonGreen and BCL6–mScarletI fusion proteins. Using time-lapse imaging, we tracked MYC and BCL6 protein expression in single B-ALL cells over 20 hours. Matching their proliferative characteristics, 'steady’ B-ALL cells expressed MYC and no detectable BCL6. However, B-ALL PDX also included 'alternating’ cells that underwent multiple transitions between MYC+ BCL6- and MYC- BCL6+ cell states. PDX from BCR-ABL1 B-ALL included approximately 30% 'alternating’ cells, whereas 'alternating’ cells exceeded 50% in B-ALL PDX with RAS-pathway lesions. Among 'alternating’ cells, transitions occurred independently from cell divisions (every 36 hours) with shorter MYC-phases of three hours and longer BCL6-phases of six hours. Integrated ChIP-seq, RNA-seq, and untargeted metabolomics revealed that MYC-high cells are much larger, activated glycolysis and protein-synthesis pathways, whereas BCL6-high cells are smaller and enriched for phosphatidylethanolamine (PtdEtn) synthesis pathways. Consistent with the essential role of PtdEtn in autophagosome biogenesis, BCL6+ B-ALL cells exhibited substantially increased autophagy (LC3B flux). To directly compare MYC and BCL6 dynamics with cell growth trajectories, we combined quantitative phase microscopy and time-lapse fluorescence imaging to measure single-cell dry mass over 8 hours. Cells in the MYC-high state accumulated biomass at twice the rate of BCL6-high cells, producing a step-by-step progression of fast (MYC) and slow (BCL6) cell mass accumulation that aligned with each MYC/BCL6 state transition. Finally, multiplex immunofluorescence of FFPE biopsies from patient samples revealed spatially segregated niches of MYC-high and BCL6-high leukemia cells reminiscent of partitioning of MYC+ centrocyte and BCL6+ centroblasts in germinal centers. To experimentally induce cell state-transitions, we engineered patient-derived B-ALL cells by HDRT with dual-reporter knockin alleles expressing C-terminal MYC–dTAG fusion proteins. In addition to the MYC-dTAG degron system, we used the BCL6 PROTAC degrader ARV-393 for acute ablation of either MYC or BCL6, which was achieved within 90 minutes. Acute dTAG-mediated degradation of MYC induced cell shrinkage and decreased cellular dry mass, whereas BCL6 degradation via PROTAC decreased autophagic flux and increased dry mass, confirming opposing roles in anabolic versus catabolic metabolism. To functionally study 'steady’ (MYC-only) and 'alternating’ (MYC/BCL6) B-ALL populations, we developed a cell-sorting strategy that enriched each of the two populations to a purity of 85% as confirmed by subsequent time-lapse imaging. Extreme limiting dilution and serial transplant experiments revealed that 'steady’ (MYC-only) largely lacked LIC-capacity. In contrast, 'alternating’ (MYC/BCL6) B-ALL cells were highly enriched for LIC and initiated fatal leukemia after short latency. Conclusion: This study uncovers previously unrecognized cell state transitions in B-ALL that are controlled by MYC- and BCL6-dependent transcriptional programs. Alternating cycles of quiescent and proliferative cell states balance anabolic and catabolic metabolism, promote drug resistance, and enhance leukemia-initiating potential.