MLL/AF9 Expression Causes Leukemia in Zebrafish

Background

Acute Leukaemia (AL) is a genetically heterogeneous disorder caused by somatic mutations and acquired chromosomal translocations. Translocations can lead to the formation of fusion genes such as the MLL/AF9 fusion, which results from the t(9;11)(p22;q23).

A better understanding the molecular pathophysiology of AML, of the mechanisms of treatment resistance, disease relapse can be achieved by developing animal models. The MLL/AF9 fusion is frequently used to model AML in mice. However, to date, no MLL/AF9 leukemia models in zebrafish have been reported.

Aim

Our aim was to establish a transgenic zebrafish leukemia model using the human MLL/AF9 fusion gene.

Methods

To generate transgenic fish, two constructs (pTol2-Runx1+23: MLL-AF9-IRES-EGFP-cmlc-GFP and pTol2-Runx1+23: MLL-AF9-IRES-mCherry-cmlc-GFP) were injected together with Tol2 transposase mRNA into one-cell stage zebrafish embryos. We used the murine Runx1+23 enhancer to direct MLL/AF9 expression to hematopoietic stem cells and EGFP or mCherry as fluorescent markers. We selected transgenic embryos 24 hours post-fertilization based on the heart marker expression (cmlc).

Results

29% (100 of 340 embryos) of the transgenics reached adulthood (6 weeks). After 6 to 24 months, 77% (77) of them developed signs of sickness. They became less active with protruding eyes and hump formation on the nose. Some started bleeding from the gills and/or showed tumor formation around the abdomen and head. Sick fish were euthanized and dissected. The autopsies showed pale and dysmorphic kidneys, pale and enlarged spleens, and in some cases white granular spots on the spleen. Histological sections revealed increased kidney, spleen, and liver cellularity with massive cellular infiltration of cells in these organs. In flow cytometry, kidney marrow cells from the transgenic fish showed a different forward scatter (FSC) and side scatter (SSC) profile compared to that of the normal zebrafish kidney marrow cells.

The transgenic F ₀ zebrafish showed increased numbers of lymphoid cells (12 fish), precursor cells (9 fish), or myeloid cells (6 fish). Peripheral blood smears showed many intermediate-sized mononuclear cells with an increased nuclear-cytoplasmic ratio, with nuclei containing dipsersed chromatin and inconspicuous nucleoli resembling blasts. There were occasional myelocytes and no mature granulocytes. These data are consistent with the development of acute leukemia in our MLL/AF9 transgenic fish. Whole-mount in situ hybridization (WISH) was performed on F ₁ embryos. RNA probes for early hematopoietic markers (gata1, scl, runx1, ikaros, cmyb, mpx, and lyz) were hybridized to 24 and 48 hpf F1 transgenic embryos. There were expression changes of these markers compared to age-matched wild-type larvae, including low expression of gata1, scl, cmyb and high expression of lyz, mpx and ikaros in the caudal hematopoietic organ.

We also performed transplantation experiments with the kidney marrow cells from diseased fish to test whether the disease was transplantable. The disease was serially transplantable into secondary and tertiary recipient fish. Transplanted fish had a significantly shorter latency to disease development of only 2 to 6 weeks. The morphological evidence and the serial transplantability of the disease proves that we have in fact succeeded in establishing an MLL/AF9-driven acute leukemia model in zebrafish.

The long latency and incomplete penetrance observed in our F ₀ MA9 zebrafish, along with a shorter latency in the transplanted fish, suggests that additional somatic mutations are required for leukemogenesis in this model. We performed whole exome sequencing to find cooperating somatic mutations and RNASeq to identify differentially expressed genes in MA9 leukemia fish.

Whole exome sequencing on six samples identified putative somatic mutations in genes such as Stat5, Cyp2j20, Ms4a17a.3, Tapbp.1 and Herc5.3, which have been reported to be mutated in human cancer. RNA-seq analysis on seven samples showed 67 differentially expressed genes with a q value < 0.05 (e.g., cxcl32b.1, myof1, ctdsp2, egr3, il2rb) and nine enriched pathways with a P-value of < 0.054 (e.g.: KRAS, TP53, MEK) in our transgenic leukemic MLL/AF9 fish.

Conclusion

Our MLL/AF9 zebrafish acute leukemia model will be a helpful tool to understand leukemia biology and enable testing of new therapeutic strategies.