Search

Despite improvements in overall survival for children with B-cell progenitor acute lymphoblastic leukemia (BCP-ALL), it remains the second-leading cause of cancer related death in children with approximately 200 deaths per year in the U.S. Thus, there remains a critical need for a definitive cure to prevent relapse for patients with BCP ALL. The accumulation of BCP ALL blasts results from the disruption of normal developmental checkpoints. One of these checkpoints, as pro-B cells transition to become pre-B cells, involves surface expression of the precursor-B-cell receptor (pre-BCR). Prior work has categorized BCP ALL into pre-BCR positive and pre-BCR negative subtypes based on the protein expression of Ig light chain and active signaling of SRC family kinases, SYK, BTK. Combining single cell analysis and machine learning, we previously identified pre-B cells with activation of pre-BCR signaling, namely CREB, 4EBP1, rpS6 and SYK, that are present at diagnosis and highly predictive of relapse. We call these relapse predictive cells. Relapse predictive cells were enriched in relapse samples, demonstrating their persistence from diagnosis to relapse and making them an actionable target to prevent relapse altogether. To better understand relapse predictive cells, we enriched pre-B cells from patients with known relapse status and performed whole transcriptome sequencing. Relapse predictive cells demonstrated significant upregulation of genes in the oxidative phosphorylation (OXPHOS), glycolysis, and reactive oxygen species (ROS) pathways compared to pre-B-like leukemia cells from patients who will not go on to relapse. Analysis of public genome-wide CRISPR screen datasets in 2 pre-BCR+ and 4 pre-BCR- cell lines found 69 essential genes uniquely present in pre-BCR+ cell lines, related to mitochondria translation, OXPHOS and TCA cycle pathway. We performed CRISPR knock down of proximal pre-BCR related tyrosine kinase SYK in pre-BCR+ (Nalm6, Kasumi-2) and pre-BCR- (697, REH, SUPB15) cell lines to understand how activated pre-BCR impacts cellular metabolism in pre-BCR+ and pre-BCR- cells. CyTOF analysis of pre-BCR signaling demonstrated effective inhibition of downstream pre-BCR pathway members in the KD cells (pSYK, pBLNK, pBTK). RNA sequencing demonstrated upregulation of mitochondrial translation and OXPHOS pathways with downregulation of hypoxia pathways in pre-BCR+ but not pre-BCR- SYK KD cells. Functional extracellular flux experiments by Seahorse confirmed pre-BCR+ SYK KD cells to have higher basal oxygen consumption rate (OCR) and lower extracellular acidification rate (ECAR) compared to wild-type pre-BCR+ cells, indicating a switch from highly glycolytic to aerobic metabolism. To determine the interplay between pre-BCR signaling and cellular metabolism at the single cell level, we performed CYTOF with a panel examining pre-BCR pathway members, developmental phenotype and metabolism in these cell lines as well as matched diagnosis-relapse patient-derived xenografts. These results indicate, in line with the RNA sequencing and Seahorse data, that inhibiting pre-BCR signaling is accompanied by inhibition of glycolysis with lower protein expression of glycolytic related enzymes HIF1A, GLUT1, PFKFB4, GAPDH, ENO1 and LDHA. Further, we observed in cells completely deficient in the ability to initiate pre-BCR signal (SYK knock out), activated p4EBP1 indicating signaling feedback from the PI3K-AKT pathway and a metabolic adaption indicating utilization of energy sources other than glucose in cells surviving SYK loss. Finally, to determine the impact of loss of pre-BCR signaling on proliferation, in vitro competition assays demonstrated SYK KD cells to be less proliferative in all the cell lines except pre-BCR- cell line 697. In vivo, SYK KO demonstrated significantly slower engraftment (median %hCD45: 84% vs 54%, P=0.009) in NSG mice and significantly longer survival time than the mice xenografted with wild-type cells (median survival 28 vs 39 days, P=0.0004). Together, our data indicate that individual BCP ALL cells with active pre-BCR signaling are associated with relapse and that these cells have a unique metabolic state that relies on active glycolysis and metabolic flexibility supporting proliferation in vitro as well as engraftment and aggressivity in vivo. Further metabolomics experiments and characterization of primary patient samples are underway. Disclosures Mullighan: Pfizer: Research Funding; Illumina: Membership on an entity's Board of Directors or advisory committees; AbbVie: Research Funding; Amgen: Current equity holder in publicly-traded company. Davis: Novartis Pharmaceuticals: Honoraria; Jazz Pharmaceuticals: Research Funding.

Glucocorticoids (GCs) remain a backbone component of therapeutic regimens for childhood B-cell acute lymphoblastic leukemia (B-ALL). GCs resistance is a strong prognostic marker of relapse making its understanding an important challenge to be addressed to improve overall patients outcome. Healthy B-cell development is characterized by checkpoints where critical regulatory signaling influences the fate of developing B-cells. These stages are vulnerable for leukemic transformation and as we previously demonstrated, the developmental and functional state of B-ALL cells are of critical importance in treatment failure. Using this developmental framework, we examined the effect of GCs on healthy and malignant B-cells to better understand mediators of GC resistance and ways to overcome it. To define the dynamics of the transcriptional networks surrounding B-cell developmental checkpoints, we performed transcriptomic analysis of sorted pre-proB, pro-B and pre-B cells from 3 healthy donors. We found a coordinated upregulation of B cell receptor (BCR) and Glucocorticoid Receptor (GCR) pathways in healthy B cells during their development. Single cell proteomic analysis of these healthy populations confirmed the coordinated expression of GCR with early B-cell phenotypic markers. Furthermore, in vitro treatment of healthy B-cells with the GC dexamethasone (dex) demonstrated cycling pro-B/pre-B cells to be the most sensitive to GC-induced cell death. Given the importance of GCs in B-ALL treatment, we investigated GCs effects and resistance in NALM6 and REH cell lines. NALM6, expressing high levels of GCR, were significantly sensitive to dex treatment in terms of cell cycle arrest and cell death. By contrast, REH cells were resistant to dex treatment as they lack the GCR. Retroviral transfection of NR3C1 (GCR gene) in REH cells (REH GCR) resulted in acquired sensitivity to dex. To explore the potential crosstalk between GCR and BCR pathways, we also treated cells with BCR signaling inhibitor, dasatinib (das), alone or in combination with dex. While both cell lines survived to treatment with das, the combined treatment increased apoptosis compared to dex alone in NALM6 cells (p=0.0179) and REH GCR cells (p=ns). Whole transcriptome sequencing of dex-resistant cells revealed upregulation of BCR downstream signaling as one of the main pathways associated with resistance. The cell line data implicated active BCR signaling as a path to GCs resistance, so we next analyzed 19 B-ALL primary samples by mass cytometry, after in vitro exposure for 48 hrs to same treatments. Across the entire cohort, dex induced a significant reduction in viability (p=0.0007), compared to vehicle. Treatment with das and dex+das also decreased cell viability compared to vehicle (p=0.0038 and p=0.0002) although was not significant to dex alone. Interestingly dex-resistant cells showed a phenotypic modulation compatible with a late pre-B phenotype with increased CD45 and CD20 expression. In addition, surviving cells showed an activation of downstream targets of BCR signaling such as pSYK, pRPS6 and pCREB that was partially blunted by dasatinib treatment. To understand whether the phenotype and signaling modulation induced by dexamethasone also occurs in patients after treatment with GCs, we analyzed minimal residual disease (MRD) cells following 8 days of treatment with GCs from 9 B-ALL patients. This analysis confirmed our previous findings with MRD cells having same late pre-B cells phenotype and signaling profile as in vitro treated cells. Finally, we tested whether targeting of pre-BCR signaling via dasatinib, could overcome resistance in vivo. We evaluated engraftment of luciferase-expressing NALM6 cells at different timepoints after tail vein injection in mice treated with vehicle, dex, das or dex + das. Bioluminescence analysis revealed a significant reduction of early and late engraftment in the dex + das group compared to vehicle-treated mice. Furthermore, mice receiving the combined treatment also experienced a significant survival advantage as assessed by log rank test (p=0.0002). Taken together these data suggest a coordinated interplay between BCR and GCR pathways in healthy and leukemic B cells. GCs-resistant leukemic cells showed a mature pre-B phenotype that is vulnerable to BCR signaling inhibition in vitro and in vivo suggesting new therapeutic options to overcome GC resistance in childhood B-ALL. Disclosures Biondi: Bluebird: Other: Advisory Board; Novartis: Honoraria; Incyte: Consultancy, Other: Advisory Board; Amgen: Honoraria; Colmmune: Honoraria. Bava: 10x Genomics: Current Employment. Davis: Jazz Pharmaceuticals: Research Funding; Novartis Pharmaceuticals: Honoraria.