Elucidating Unifying Mechanisms for Inherited Predisposition to Childhood Acute Lymphoblastic Leukemia

The development of early interception or preventative approaches for childhood B-cell acute lymphoblastic leukemia (B-ALL) is dependent on a deeper understanding of the mechanisms predisposing children to acquire this malignancy. While tremendous progress has been made to define somatic drivers and their underlying mechanisms in B-ALL, a significant heritable component for this disease is increasingly being recognized. Insights into some of this heritable risk have arisen through genome-wide association studies for common risk variants, and familial studies have defined high-impact rare risk variants. Despite the robust genetic observations made, however, it is unknown whether there are general unifying predisposition mechanisms. To address this limitation, here, we have performed systematic perturbations with rich single-cell genomic readouts to gain insights into underlying mechanisms.

To systematically perturb and study predisposition mechanisms, we developed a robust in vitro semi-synchronous B-cell differentiation system from umbilical cord blood-derived human hematopoietic stem and progenitor cells (HSPCs). We have comprehensively characterized the precise cellular composition of this differentiation platform by flow cytometry, B-cell receptor profiling, and single-cell RNA sequencing and have confirmed the presence of key intermediate progenitors and precursors, including pro- and pre-B cells, after 3-4 weeks of culture. Additionally, we have found this platform to be amenable to highly efficient CRISPR/Cas9 genome editing in HSPCs, thereby enabling us to directly investigate genome-edited cells that are undergoing differentiation through the sensitive and transient states in which genetic predisposition might act.

To prosecute the impact of genes harboring high-penetrance germline B-ALL predisposition variants on early B-cell development, we performed CRISPR/Cas9 editing to recreate known high-impact germline loss-of-function B-ALL risk variants in CDKN2A, ETV6, IKZF1, NBN, PAX5, SH2B3, TCF3, TP53 and USP9X. Efficient gene editing was achieved by targeting these genes in human HSPCs using two validated single guide RNAs (sgRNAs) per target, as well as control sgRNAs targeting the AAVS1 safe-harbor site. Following four weeks of in vitro B-cell culture, we found distinct differentiation blocks in early B-cell development for cells edited for ETV6, IKZF1, PAX5 and TCF3 through flow phenotypic analyses. Next, to investigate how these perturbations impact B-cell differentiation at higher resolution, we performed single-cell RNA sequencing. We obtained transcriptomic data from a total of 47,748 cells, which enabled us to assess sgRNA depletion or enrichment across specific cell states, as well as alterations within a specific differentiation stage. Through these computational analyses of this rich functional dataset, we observe consistent stalling of B-cell development for all relevant perturbations, albeit at variable progenitor states. Importantly, the observed stalling is nearly always occurring in pro-B and pre-B cells, at cell states where RAG1 and RAG2 are robustly expressed, suggesting potential mechanisms through which the predisposition can arise. Further analyses are being performed to define how such stalling at stages with high-level RAG recombinase expression can predispose to the acquisition of B-ALL - a disease known to be driven by illegitimate RAG-mediated recombination.

In summary, we have developed an in vitro platform for the systematic recreation of high-impact germline risk variations for B-ALL, which has enabled us to elucidate recurrent mechanisms of stalling and delayed differentiation, and which suggest potential unifying mechanisms by which B-ALL predisposition can arise.

Sankaran:Ensoma: Consultancy, Honoraria.