The DLX homeodomain genes are part of the Drosophila distal-less family, originally identified in the forebrain of the developing mouse embryo. DLX1 gene is expressed also in the hematopoietic cells. Our previous data showed that patients with FLT3/ITD (internal tandem duplication) mutation representing about 35% of all acute myeloid leukemia (AML) cases have higher expression of DLX1 compared to non-FLT3/ITD AML patients. Further, FLT3 signaling was described to regulate DLX1 gene expression.

In the current study we found an association between DLX1 expression level and overall survival using GEP data from TCGA database within the group of FLT3/ITD-positive patients. Worse overall survival was linked to lower expression of DLX1 (p=0.003, n=46) even though generally this group of patients is characterized by their higher DLX1 level. This association was also observed when all AML patients were analyzed (p=0.01, n=197). Since the role of DLX1 in the leukemogenic process is not yet described we aimed to characterize the phenotype of leukemic cells with different expression levels of DLX1 gene.

We designed two functional shRNAs (pGhU6-sh1_DLX1 and pGhU6-sh2_DLX1, referred as sh1 and sh2) to downregulate DLX1 expression in MV4;11 (FLT3/ITD positive) leukemic cells, which present high endogenous level of DLX1, and a non-silenced control (pGhU6-NSC). In vitro studies showed that cells with silenced DLX1 were arrested in G0 phase of cell cycle (35%±4.8 (NSC) to 67.5%±2.2 (sh1; p ˂0.01) and 65.5%±2.5 (sh2; p ˂0.01); flow cytometry - Pyronin Y/ Hoechst 33342 staining) and had lower proliferative activity (trypan blue over the period of 10 passages). Moreover, cells with silenced DLX1 were less apoptotic (Annexin V/PI staining).

Next, we studied the impact of DLX1 downregulation on leukemic cell infiltration in vivo using sub-lethally irradiated NSG (NOD SCID gamma) mice. We injected 1x106 silenced or non-silenced MV4;11 cells via tail vein (n=20, 6 mice/group, 2 control mice). After two weeks we measured the absolute number of MV4;11 cells by flow cytometry (CD33+/GFP+/DAPI-) in bone marrow (BM) and spleen (SP) of recipient mice. Absolute cell counts of leukemic cells with silenced DLX1 were 1.26x106 (sh1; p=0.015) and 2.89x106 (sh2; p˂0.0001) vs NSC cells 0.52x106 in BM and 1.59x106 (p=0.036) and 14.2x106 (p=0.025) vs 0.34x106 in SP. To ensure that the increased cell numbers were not the result of enhanced homing of DLX1 silenced cells, we performed homing experiments. Control and DLX1 silenced cells were transplanted and the number of cells in BM and SP was determined 16 hours after transplantation. We observed no differences between the studied groups, indicating that DLX1 silencing does not affect the homing ability of MV4;11 cells. Moreover, we analyzed cell cycle in leukemic cells isolated from recipient mice three weeks after transplantation. In agreement with our in vitro results, leukemic cells with silenced DLX1 had higher percentage of cells arrested in G0 phase (48.6%±6.3 (sh1; p ˂0.001) and 80.9%±9.3 (sh2; p ˂0.001) vs 26.4%±4.8 (NSC) in SP). Overall, mice with silenced DLX1 presented worse fitness and bigger splenomegaly.

Further, we investigated the signaling pathways which could lead to G0 arrest. Since DLX1 inhibits TGF-β pathway through direct interaction with SMAD4, a key downstream effector of TGF-β/BMP signaling, we studied the changes in target genes in our model cell lines. Expression of CUTL-1 (sh1 - 1.5-fold, sh2 - 2-fold change to NSC), PAI-1 (3.5-fold, 4.2-fold) and CDKN1C (1.3-fold, 2.3-fold) were significantly increased in DLX1 silenced cells. PAI-1 was shown to induce replicative senescence and CDKN1C is an inhibitor of cell cycle progression. While both targets disturb cell cycle and could be responsible for the phenotype we have observed, this hypothesis needs to be elucidated in future experiments.

Altogether, our data demonstrate that dysregulation of DLX1 gene in leukemic cells changes the cell phenotype. Lower level of DLX1 gene leads to arrest in G0 phase which in vitro slows down the proliferation whereas in vivo it allows the cells to persist in spleen and BM. We hypothesize that the DLX1 silenced cells become more resistant to external effects which could then be reflected as a reduced survival observed in patients with low levels of DLX1.

Supported by P304/12/2214.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution