Introduction: Children with Down syndrome (DS) are 10-20 times more likely than children without DS to develop acute lymphoblastic leukemia (ALL), and they demonstrate a distinctive spectrum of genetic alterations. Approximately 50% of DS-ALL cases demonstrate CRLF2 rearrangements (CRLF2-R), an approximately 10-fold higher frequency than in non-DS ALL. We sought to identify the functional basis for the increased incidence of ALL, and specifically CRLF2-R ALL, in children with DS.

Methods: We created retroviral vectors which induce overexpression of CRLF2 and green fluorescent protein (GFP) for transduction into bone marrow (BM) cells isolated from the Dp16(1)Yey (Dp16) mouse model of DS, which is trisomic for the approximately 115 human chromosome 21 gene orthologs present on mouse chromosome 16. Transduced BM cells from Dp16 and wild-type (WT) control mice were co-cultured with OP9 stromal cells for one week to promote B-lymphoid lineage development, and then characterized by flow cytometric Hardy fraction analysis, or grown in B-lymphoid-promoting methylcellulose medium for colony growth assays.

Results: We achieved efficient transduction (80-95%) of Dp16 and WT BM enriched for hematopoietic stem cells (HSCs) with CRLF2-GFP+ and control GFP+ viruses. Following OP9 co-culture, transduced HSCs were characterized by Hardy fraction analysis. CRLF2-GFP+ Dp16 lymphoid cells demonstrated significantly higher percentages of immature Fraction A (pre-pro-B) cells compared with GFP+ Dp16 cells (39.9% vs 15.7%, p=0.004, Fig. 1A). This CRLF2-GFP-induced immature immunophenotype was more pronounced in Dp16 versus WT HSCs, with a significantly higher percentage of Fraction A cells (39.9% in Dp16 vs 24.0% in WT, p=0.0002) and a significantly lower percentage of more mature Fraction B (pro-B) cells (24.3% in Dp16 vs 49.1% in WT, p=0.02, Fig. 1A,B). In methylcellulose colony assays, CRLF2-GFP+ Dp16 cells yielded a 36-fold increase in B cell colonies compared to GFP+ Dp16 cells (Fig 1C). Again, the effect of CRLF2 transduction was magnified in the Dp16 versus WT background. CRLF2-GFP+ WT cells demonstrated only a 2.9-fold increase in B cell colonies (Fig 1C).

Conclusions: Here we demonstrate that CRLF2 overexpression results in a more immature B-lineage immunophenotype and increased lymphoid colony growth in vitro, and that these effects are significantly greater in a murine DS versus WT genetic background. Experiments to investigate the pathways involved and to evaluate these effects in vivo are ongoing. This work provides functional evidence of the enhanced leukemogenicity of CRLF2 overexpression in DS-ALL, and creates a tractable model system for additional future genetic studies.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution