Insights into the cellular origin and etiology of the infant pro-B acute lymphoblastic leukemia with MLL-AF4 rearrangement

A human ESC model for MLL-AF4 leukemic fusion gene reveals an impaired early hematopoietic-endothelial specification

The MLL-AF4 fusion gene is a hallmark genomic aberration in high-risk acute lymphoblastic leukemia in infants. Although it is well established that MLL-AF4 arises prenatally during human development, its effects on hematopoietic development in utero remain unexplored. We have created a human-specific cellular system to study early hemato-endothelial development in MLL-AF4-expressing human embryonic stem cells (hESCs). Functional studies, clonal analysis and gene expression profiling reveal that expression of MLL-AF4 in hESCs has a phenotypic, functional and gene expression impact. MLL-AF4 acts as a global transcriptional activator and a positive regulator of homeobox gene expression in hESCs. Functionally, MLL-AF4 enhances the specification of hemogenic precursors from hESCs but strongly impairs further hematopoietic commitment in favor of an endothelial cell fate. MLL-AF4 hESCs are transcriptionally primed to differentiate towards hemogenic precursors prone to endothelial maturation, as reflected by the marked upregulation of master genes associated to vascular-endothelial functions and early hematopoiesis. Furthermore, we report that MLL-AF4 expression is not sufficient to transform hESC-derived hematopoietic cells. This work illustrates how hESCs may provide unique insights into human development and further our understanding of how leukemic fusion genes, known to arise prenatally, regulate human embryonic hematopoietic specification.

Challenges in the use of NG2 antigen as a marker to predict MLL rearrangements in multi-center studies

Rearrangements in MLL (MLL-r) are common within very young children with leukemia and affect the prognosis and treatment. Previous studies have suggested the use of the NG2 molecule as a marker for MLL-r but these studies were performed using a small number of infants. We analyzed 148 patients (all less than 24 months, 86 less than 12 months) from various centers in Brazil to determine the predictive power of NG2 within that cohort. We show that NG2 can be used for MLL-r prediction; however, proper staff training and standardized sampling procedures are essential when receiving samples from multiple centers as the accuracy of the prediction varies greatly on a per center basis.