RNA Binding Protein CELF2 Mutation Cooperates with MLL-AF9 in Driving Acute Myeloid Leukemia Development

Acute myeloid leukemia (AML) is an aggressive form of hematologic malignancies, caused by the accumulation of immature leukemic blasts in the hematopoietic system. Current chemotherapy regimens remain ineffective as assessed by low remission rate and 5-year-survival rate. MLL-AF9 mutations occur in approximately 10% of all acute leukemia patients, and are associated with poor therapy response and prognosis. Deregulated expression of ELAV-like family protein 2 (CELF2) has been identified in AML patients with MLL-AF9. As an RNA binding protein, CELF2 has been shown to regulate RNA splicing and embryonic hematopoietic development. However, the role of CELF2 in AML progression is still largely unknown.

In this study, we first analyzed genetic data of 160 AML cases in the Cancer Genome Atlas in which CELF2 deletion account for 10% of the total cases. The survival time of patients with CELF2 deletion was much shorter than that of the others (P < 0.001). Interestingly, the expression of CELF2 was significantly decreased in different sub-type of acute myeloid leukemia cells, and the lowest expression of CELF2 was seen in MLL-AF9 (MA9) AML cells, when compared with that of normal myeloid cells. These results suggested that CELF2 deletion/mutation may play an important regulatory role in MA9-induced AML progression.

To evaluate the potential role of Celf2 in the hematopoietic malignant transformation, we generated hematopoietic specific Celf2-deficient mice (Vav1-Cre;Celf2^fl/fl) and found that Celf2-deficient hematopoietic stem cells (HSCs) over-expanded upon transplantation indicating a role of Celf2 in HSC proliferation. Consequently, the percentage of Celf2-KO HSC-derived myeloid cells was significantly increased in peripheral blood and bone marrow of recipient mice when compared with that of control mice. Celf2-KO Lin- cells were then infected by MA9-containing retrovirus and transplanted into lethally irradiated mice to generate Celf2-KO+MA9 mice. All Celf2-KO+MA9 mice succumbed to AML in 35 days, and their survival time is shorter than that of Celf2-WT+MA9 mice. The Celf2-KO+MA9 mice showed more aggressive phenotypes of AML than Celf2-WT+MA9 mice, including higher counts of WBCs, neutrophils and lymphocytes in peripheral blood, as well as higher extramedullar infiltration in spleen, liver and lung. To quantify the functional LSCs in Celf2-KO+MA9 mice and control MA9 mice, extreme limiting dilution transplantation assay was performed. The penetrance rate of leukemia in mice that received 1x10³Celf2-KO+MA9 cells was significantly higher than that of control mice (100% versus 50%). The frequency of functional LSCs was significantly increased in Celf2-KO+MA9 mice as compared to that of the control, supporting a role of loss of Celf2 in promoting leukemia blast self-renewal.

To investigate the underlying molecular mechanisms of Celf2 deletion in accelerating MA9 AML progression, we performed RNAseq to analyze the transcriptome programing changes associated with Celf2 in isogenic Celf2-WT+MA9 or Celf2-KO+MA9 AML cells. Consistent with severe extramedullar infiltration of leukemia cells in Celf2-KO+MA9 mice, gene set enrichment analysis showed up-regulated leukocyte activation and migration in Celf2-KO+MA9 leukemia cells. We next analyzed the downstream signaling pathways targeted by Celf2, and found that the phosphorylation of S6 was increased in Celf2-KO+MA9 cells, suggesting that Celf2 deletion might promote leukemia progression by activating mTORC1 signaling pathway. To confirm this, we treated Celf2-KO+MA9 mice with MA9 inhibitor (EPZ5676) and mTORC1 inhibitor (Rapamycin), and found a collaborative response to the treatment with a significant attenuation in leukemia burden in vivo.

In conclusion, we found that CELF2 low expression is associated with AML patients with MA9 mutation. Celf2 deletion in mice induced myeloid-biased differentiation of HSCs. Notably, Celf2 deficiency promoted MA9-induced AML progression by activating mTORC1 signaling pathway independent of MA9 signaling pathway. Combinational therapy with MA9 inhibitors and mTORC1 inhibitors could significantly decrease the Celf2-KO+MA9 leukemia progression in vivo, indicating that combinational usage the drugs may have synergistic benefit for MA9 AML patients with low CELF2 expression.