Distinctive functions of the polycomb group protein BMI-1 in hematopoiesis and leukemogenesis

Abstract

Bmi-1 maintains stem cell population in hematopoietic system for replenishment of progenitors and mature cells. It has been shown that Bmi-1 prevents stem cell exhaustion through suppression of cell cycle regulators 〖p16〗^INK4a/ 〖p19〗^Arf and cell differentiation. However, it remains unclear how Bmi-1 maintains self-renewal of hematopoietic stem cells (HSC). To dissect the underlying mechanisms of Bmi-1 in sustaining HSC self-renewal capacity, transcriptome analysis of Bmi-1 knockdown or over-expressing hematopoietic stem and progenitor cells (HSPC) was performed. RNA-Sequencing analysis demonstrated that Bmi-1 de-regulated genes in canonical Wnt signaling, stem-cell quiescence and Tnf/Gzmb-mediated apoptotic signaling in HSPC. Moreover, it was found that Bmi-1 over-expression mildly activated canonical Wnt signaling and de-regulated a panel of Wnt-associated genes Cdkn1a, n-Myc, Fn-1 and Hoxb4 in HSPC. ChIP analysis validated that Bmi-1 binds to the promoter region of Wnt negative regulator Amer2. It suggests that Bmi-1 activates canonical Wnt signaling through suppression of Amer2 in HSPC. More importantly, ChIP analysis validated that Bmi-1 binds to its own promoter region, suggesting that endogenous Bmi-1 expression is maintained through self-regulatory negative feedback mechanism to prevent excessive Wnt signaling activation in HSPC. In view of this, a model of Bmi-1-mediated suppression of Amer2-dependent β-catenin degradation in HSC is proposed.

On the other hand, BMI-1-mediated 〖p16〗^INK4A leukemogenic pathway has been proposed in human leukemias. However, clinical studies revealed that high BMI-1 expression may not be well-correlated with low 〖p16〗^INK4A expression. Importantly, leukemogenic factors such as MLL fusion proteins can regulate 〖p16〗^INK4A expression such that the regulation of 〖p16〗^INK4A is not dependent on BMI-1. It is therefore unclear how BMI-1 is involved in the process of leukemogenesis. Although it has been demonstrated that high BMI-1 expression was correlated with disease development in human leukemias, recent studies revealed a tumor suppressive role of BMI-1 in cancers, questioning the role of BMI-1 in leukemogenesis. In order to find out the regulatory mechanism of BMI-1 in leukemogenesis, BMI-1 was overexpressed in a panel of leukemia cell lines, including HL-60, MonoMac-6, MV4-11, SEM and Nalm-6. My results demonstrated that BMI-1 over-expression suppresses JAK-STAT signaling through up-regulation of SOCS genes. Moreover, it was found that BMI-1 over-expression suppresses IL7 signaling pathway in SEM leukemia cells, leading to reduced expression of cell survival genes, including PAX5, MCL-1, BCL-2 and BCL-XL. These findings suggest that BMI-1 has a tumor suppressive function in leukemogenesis. In summary, the discoveries of Bmi-1-mediated canonical Wnt signaling and the suppressive role of BMI-1 in JAK-STAT leukemogenic signaling would provide new insights on the understanding of stem cell and leukemia biology, building a foundation for improvement of clinical applications.