Development and progression of leukemia requires interaction of leukemia-initiating cells with their bone marrow niches. The niches serve as the nursery and shelter for the leukemic cells, which can result in drug resistance, disease recurrence, and minimal residual disease, the most important causes for the death of patients with leukemia. Therefore, obliteration of the interaction between the leukemic cells and their niches is of utmost importance in eradicating leukemic cells during therapy to cure the disease. However, little is currently known of the molecular mechanisms underlying the interaction of the two types of cells. Sox4, a SRY-related HMG-box containing transcription factor that is vital during development, plays an important role in leukemia. Published mouse studies demonstrated that increased expression of Sox4 was associated with leukemogenesis. We determined the expression levels of Sox4 by real-time RT-PCR in 100 human leukemic samples and found high levels of expression in B- and T-ALL, but not in AML, CML, CLL, Sezary syndrome, or T cell prolymphocytic leukemia. In accordance, 7 of the 8 ALL cell lines (the exception was 697) we tested showed high expression levels of Sox4, but AML cell lines, normal mature B cells, T cells, and bone marrow CD34+ cells had low levels of expression. Since the majority of clinical B-ALL cases correspond to the pre-B cell stage, we investigated the role of Sox4 in a pre-B cell line (Nalm6) by lentivirus-mediated RNAi. Remarkably, knockdown of Sox4 in Nalm6 cells caused 70% reduction in the formation of leukemic cell clusters under the monolayer of co-cultured M2-10B4 bone marrow stromal cells, a phenomenon known as pseudo-emperipolesis. Similar results were obtained with ex vivo cultured bone marrow cells from conditional Sox4 knockout mice that displayed B cell developmental arrest at the transition from pro-B to pre-B cell stage and an absence of pre-B cells. These findings suggested that Sox4 is required for the interaction of the developing B cells or leukemic cells with bone marrow stromal cells, a component of the bone marrow niche. Since CXCR4/SDF1-mediated “homing” is known to be required for pseudo-emperipolesis, we tested the effect of Sox4 on Nalm6 cell migration toward SDF1 gradient and found that Sox4 did not affect the migration, suggesting that Sox4 is not acting through “homing”. Instead, our data indicated that the role of Sox4 in the interaction of leukemic cells with stromal cells is most likely mediated by its ability in enhancing the adhesion of the leukemic cells because we found that lentivirus-medicated overexpression of Sox4 in the 697 B cell line caused the suspension cells to display a spindle and adhesive morphology. In addition, 21% of the putative Sox4 downstream genes that we identified by multiple sets of gene expression microarray experiments are known to be involved in cell adhesion. Moreover, we found that the changes in gene expression profile of leukemic cells upon Sox4 knockdown or overexpression significantly overlap with the changes in response to the presence of bone marrow stromal cells in co-culture, indicating that Sox4 pathways are involved in leukemic cell response to stromal cell signaling. Based on these findings we hypothesize that deletion of Sox4 abolishes the interaction between the developing lymphocytes and their niches during lymphopoiesis. Conversely, overexpression of Sox4 may enforce these cells to over-interact with the niches so that they are overexposed to local growth factor stimuli. If superimposed with other genetic and/or epigenetic changes in the developing lymphocytes, such over-interaction may result in the development of leukemia. In case of established leukemia, such over-interaction may lead to the enhanced protection of leukemic cells by their niches. Therefore, the role of Sox4 in the interaction of developing lymphocytes or leukemic cells with their niches is like “rooting into the soil” of a growing tree, abbreviated as “rooting”.
Read full abstract