Picalm/MLLT10 Expression in Zebrafish Is Associated with an Increase of Hematopoietic Stem Cells and Leads to Leukemia

The PICALM/ MLLT10 (C/A)fusion gene is the result of the (10;11)(p13;q14) translocation, and is found in about 1% of all AML patients, 30% of T-cell ALL with a TCR γ/∂ rearrangement and in some patients with lymphoma. AML patients with a t(10;11)(p13;q14) have poor long-term survival. Mouse models of t(10;11)(p13;q14) show that C/A and the C/A minimal fusion (CA/MF) are strong oncogenic drivers, and that C/A must be expressed in the hematopoietic stem cells (HSCs) to cause leukemia. Patient samples and mouse models of C/A leukemia show that these leukemias harbor additional, cooperating mutations.

Zebrafish (ZF) have become an attractive animal model for studying hematological disorders. ZF can be easily genetically modified and kept in large numbers. Although T cell ALL models have been successful in ZF, AML models have not been as successful. Most ZF AML models fail to recapitulate the phenotype, especially transplantability.

Here we report a C/A transgenic zebrafish AML model (R:CA/MF). We used the murine Runx1 promoter to direct CA/MF expression to zebrafish HSCs using Tol2-mediated transgenesis. We detected the expression of the CA/MF transcript using RT-PCR in 3 dpf (days post fertilisation) heart marker positive transgenic F0 embryos and the expression of FLAG-tagged CA/MF protein with whole mount in situ immuno histochemistry in stable transgenic R:CA/MF embryos at 3 dpf. We performed whole mount in situ hybridisation (WMISH) and qPCR for hematopoietic genes to examine the impact of CA/MF on zebrafish hematopoiesis. Based on qPCR analysis, we observed that the expression of pu1 (p value <0.0120) and gata1 (p value <0.0004) was significantly decreased in 3 dpf R:CA/MF embryos compared to WT, whereas the expression of mpx at 2 dpf (p value <0.0132) and 3 dpf (p value <0.0378) was significantly increased. Our WMISH analysis showed that the expression of lyz was significantly decreased in R:CA/MF compared to WT embryos at 2 dpf.

To investigate the effect of CA/MF expression on HSCs, we injected the R:CA/MF construct into fertilised oocytes of R:GFP;kdrl:RFP fish (R:GFP marks HSCs and Kdrl:RFP marks blood vessels). The results from intravital confocal microscopy showed that the 3 dpf R:GFP larvae that had been injected with the R:CA/MF vector had about a 2.5 fold increase (p value <0.0001) of the GFP expressing cells (HSCs) in both the caudal hematopoietic tissue and circulation compared to control (R:Gal4) and un-injected larvae. These results were also confirmed by flow cytometry.

CA/MF transgenic zebrafish developed an aggressive hematological malignancy with a latency of about 1 year and 60 to 70% penetrance. 68% of F1 R:CA/MF (15/22), 64% of F2 R:CA/MF (45/70), 62% of F3 R:CA/MF (22/35) and 62% of F4 R:CA/MF (25/40) adult ZF were euthanised or found dead. The median survival for F1 CA/MF transgenic ZF was 65.6 weeks and 69, 78.4, and 79.1 weeks for F2, F3, and F4, respectively. Histology analysis of R:CA/MF leukemic fish showed hypercellularity in kidney, spleen, heart, gills and eyes. Myeloperoxidase (MPO) staining showed that the number of MPO positive cells in kidney marrow was about 50% higher (p value <0.0001) in R:CA/MF compared to WT. Peripheral blood film analysis of R:CA/MF showed the presence of blast-like cells. In addition, R:CA/MF malignancies were associated with an expansion of either the precursor compartment (N=20/31), the myeloid compartment (N=8/31) or the lymphoid compartment (N=3/31) on forward/side scatter flow cytometry analysis. Transplantation assays showed that R:CA/MF leukemias were serially transplantable. The latency to leukemia development was about six weeks in the first, second, and third recipient fish, which was much shorter than the latency for the primary R:CA/MF leukemias. Our preliminary whole-exome sequencing (WES) analysis identified somatically mutated genes in CA/MF zebrafish leukemia with some of these genes having human orthologues that are known to be somatically mutated in human leukemia suggesting that these are cooperating driver mutations. Preliminary RNA-Seq experiments revealed that genes in the KRAS and MTOR pathways had increased expression in the CA/MF ZF leukemia. R:CA/MF leukemia model is a valuable model for understanding the molecular mechanisms of malignant cell transformation of C/A. It allows us to investigate the effects of C/A on HSCs in vivo and to study the fully developed leukemia via serial transplantation, WES, and RNA-Seq.