A Genome-Wide CRISPR Screen Implicates MYC Dysregulation in TCF3-PBX1 B-ALL

Introduction: Acute lymphoblastic leukaemia (ALL) is the most common paediatric cancer, of which, precursor B-cell ALL (B-ALL) accounts for approximately 80% of diagnoses. B-ALL is a heterogeneous disease, with patients characterised and risk stratified according to their cytogenetic profile. TCF3-PBX1 B-ALL was associated with a poor prognosis, but on current therapies outcome has greatly improved. However, approximately 10% of these patients relapse with their disease and at this stage have a dismal prognosis (Moorman et. al. Lancet Oncology 2010). Thus, this subset of patients pose a clinical challenge, and further understanding of disease mechanisms in relapsed TCF3-PBX1 is required to aid the discovery of novel targets for therapy.

Methods: Clustered regularly-interspaced short palindromic repeats (CRISPR) technology was utilised as a whole genome CRISPR knockout (GeCKO) screen (Sanjana et. al. Nature Methods 2014) for genome-wide identification of candidate oncogene and tumour suppressor genes (TSGs) in B-ALL cell lines, including 697 (relapsed TCF3-PBX1), REH (ETV6-RUNX1), NALM16 (hypodiploid) and HAL-01 (TCF3-HLF). The GeCKO screens were performed at 300-fold library coverage, with transduced and selected cells harvested at day 0, weeks 2, 4, 6 and 8. DNA was sequenced across the integrated sgRNA region. The abundance of sgRNA constructs was analysed over time using the model-based analysis of genome-wide CRISPR-Cas9 knockout (MAGeCK) program (Li et. al. Genome Biology 2014) to identify candidate oncogenes and TSGs. Pathway analysis was performed for the identification of significantly dysregulated pathways in TCF3-PBX1 using the 697 data set and the MaGeCK Gene Set Enrichment Analysis pathway program.

Results: Whole genome CRISPR screening in the 697 cell line successfully identified 2213 candidate oncogenes (false discovery rate (FDR) <0.05, p value <0.012) and 5 candidate TSGs (FDR <0.3, p value < 6.7x10^-5). Amongst the significant candidate oncogenes was MYC (FDR = 3.0x10^-6, p value = 2.3x10^-8, rank 16) and MaGeCK Gene Set Enrichment pathway analysis using the Molecular Signatures Database (MSigDB) Hallmarks data set, identified HALLMARK_MYC_TARGETS_V1 and HALLMARK_MYC_TARGETS_V2 as the two most significant negatively regulated pathways within the knockout screen (FDR of 1.79x10^-55 and 9.876x10^-21, respectively). These data indicate a role for MYC dysregulation in TCF3-PBX1 B-ALL. Additionally, the identification of the intriguing TSG candidates CREBBP (FDR = 0.059, p value = 5.5x10^-6, rank 2), MLXIP (FDR = 0.24, p value = 3.4x10^-5, rank 3), HIF1A (FDR = 0.29, p value = 6.7x10^-5, rank 4) and ARNT (FDR = 0.02, p value = 1x10^-6, rank 1) further highlighting the importance of MYC signalling in TCF3-PBX1 B-ALL. HIF1A inhibits MYC by direct interaction, induction of the repressor MXI1 and by coordinating the degradation of MYC by the proteasome (Zhang et. al. Cancer Cell 2007, Corn et. al. Cancer Biol. Ther. 2005, Gordan et. al. Cancer Cell 2007). ARNT dimerises with HIF1A and is responsible for the recruitment of transcriptional coactivators for the transcriptional output of HIF1A (Partch et. al. Proc Natl Acad Sci USA 2011), while CREBBP has been reported to have a role in MYC inhibition in G1, preventing the progression from G1 to S phase (Rajabi et. al. J. Biol. Chem. 2005), in addition to interaction with the candidate TSG, HIF1A (Ema et. al. EMBO J. 1999, Bhattacharya et. al. Genes Dev. 1999, Park et. al. Mol. Pharmacol. 2008). Despite the role of MLXIP in glycolysis in B-ALL (Wernicke et. al. Leuk Res. 2012), and MLXIP knockout being synthetic lethal in MYC-overexpressing neuroblastoma cell lines (Carroll et. al. Cancer Cell 2015), CRISPR mediated knockout of MLXIP in the 697 TCF3-PBX1 cell line promoted growth. This may be because knockout of MLXIP reduces the number of MLXIP-MLX heterodimers to compete with MYC-MAX heterodimers at E-box sequences, permitting transcriptional activation of MYC-target genes (Carroll et. al. Cancer Cell 2015). This study has identified a number of potential mechanisms by which MYC deregulation can occur in TCF3-PBX1 B-ALL, highlighting the essential role of MYC in disease maintenance. MYC inhibition offers a potential avenue for targeted therapy in relapsed TCF3-PBX1 B-ALL, which warrants further investigation.