KDM2b/JHDM1b, an H3K36me2-specific demethylase, is required for initiation and maintenance of acute myeloid leukemia

Abstract

The histone H3 lysine 36 dimethyl-specific demethylase KDM2b/JHDM1b, which is highly expressed in various human leukemias, was previously found to be important in regulating cell proliferation and cellular senescence. However, its functions in leukemia development and maintenance are unclear. Here, we demonstrate that ectopic expression of Kdm2b/Jhdm1b is sufficient to transform hematopoietic progenitors. Conversely, depletion of Kdm2b/Jhdm1b in hematopoietic progenitors significantly impairs Hoxa9/Meis1-induced leukemic transformation. In leukemic stem cells, knockdown of Kdm2b/Jhdm1b impairs their self-renewing capability in vitro and in vivo. The functions of Kdm2b/Jhdm1b are mediated by its silencing of p15(Ink4b) expression through active demethylation of histone H3 lysine 36 dimethyl. Thus, our study suggests that Kdm2b/Jhdm1b functions as an oncogene and plays a critical role in leukemia development and maintenance.

Figure 1

Expression of Kdm2b/Jhdm1b is up-regulated in leukemic stem cells. (A) KDM2B/JHDM1B is overexpressed at the RNA level in human AML, T-cell acute lymphoid leukemia (T-ALL), and B-cell acute lymphoid leukemia (B-ALL) compared with their normal controls. Data are derived from Oncomine. (B) Kdm2b/Jhdm1b is overexpressed in Hoxa9/Meis1-induced AML samples compared with that of normal BM. Relative mRNA levels are measured by RT-qPCR and normalized to Gapdh level. The level in normal BM is arbitrarily set to 1. (C) Separation of Hoxa9/Meis1-transformed murine AML cells into c-kit^high and c-kit^low populations by FACS. These 2 cell populations are used for methylcellulose colony formation assays. (D) The c-kit^high cell population has higher colony formation capability compared with the c-kit^low population. (E) RT-qPCR analysis shows that the c-kit^high cell population expresses a higher level of Kdm2b/Jhdm1b than the c-kit^low cell population. Relative mRNA levels are measured by RT-qPCR and normalized to Gapdh level. The level of c-kit^low population is arbitrarily set to 1. FSC indicates forward scatter.

Figure 2

Enforced expression of Kdm2b/Jhdm1b is sufficient for BM transformation. (A) Enforced expression of Kdm2b/Jhdm1b is capable of immortalizing BM cells in vitro. Shown are colony numbers of c-kit⁺ HPCs transduced with lentiviral vectors expressing GFP, Kdm2b/Jhdm1b, or Hoxa9-Meis1 at each round of methylcellulose replating. (B) Growth curve of c-kit⁺ hematopoietic progenitors transduced with lentiviral vectors expressing GFP, Kdm2b/Jhdm1b, or Hoxa9-Meis1 in suspension culture. (C) FACS analysis of the genetically modified Ly5.2⁺ donor cells, LacZ and Jhdm1b overexpression (Jhdm1b OE) in the peripheral blood of recipients. The results show that the percentage of Kdm2b/Jhdm1b-overexpressing Ly5.2⁺ cells increases gradually after transplantation. (D) Methylcellulose replating assay shows demonstrates that Kdm2b/Jhdm1b-overexpressing BM cells isolated from recipient mice can form increased numbers of colonies. In contrast, mock-transplanted BM cells fail to grow continuously in the methylcellulose replating assay. Colony numbers for each round of replating are shown. (E) Flow cytometric analysis shows that Kdm2b/Jhdm1b-overexpressed colonies express a high level of c-kit and low levels of myeloid lineage markers Mac-1/Gr-1. (F) Splenomegaly is observed in mice that received a transplant with Kdm2b/Jhdm1b-overexpressing BM cells. Shown is a representative picture of spleens harvested from mice 6 weeks after transplantation of BM cells transduced with lentiviral vectors expressing LacZ or Kdm2b/Jhdm1b. Bar size represents 1.0 cm. (G) May-Grünwald/Giemsa staining showing typical leukoblasts in the BM of mice that received a transplant with Kdm2b/Jhdm1b-overexpressing BM cells. Bar size represents 10 μm. (H) Survival curve shows mice that received a transplant with mock cells survived ≥ 210 days after transplantation, whereas most mice that received a transplant with Kdm2b/Jhdm1b-overexpressing cells died within 120-180 days after transplantation.

Figure 3

Kdm2b/Jhdm1b is required for Hoxa9-Meis1-induced leukemic transformation in vitro. (A) Flow chart of experimental procedure. To examine the role of Kdm2b/Jhdm1b in Hoxa9-Meis1-induced leukemic transformation in vitro, c-kit⁺ hematopoietic progenitors were isolated from E14.5 fetal liver (FL) and transduced with lentiviral (LV) vectors expressing various combinations of proteins and shRNAs. Transduced cells were then plated on methylcellulose to evaluate the effect of Kdm2b/Jhdm1b KD on colony formation and replating capacity. (B) KD of Kdm2b/Jhdm1b in Hoxa9-Meis1-induced leukemic cells impairs their methylcellulose colony replating capacity. Colony numbers for each round of replating are shown. (C) Growth curves indicate KD of Jhdm1b in Hoxa9/Meis1/GFP (J1bKD-HMG)–transformed leukemic cells impairs cell proliferation compared with Hoxa9/MeisI/GFP-transformed cells with control KD (CKD-HMG). Transformed cell colonies were picked after the third round of methylcellulose replating and cultured in liquid medium. (D) KD of Kdm2b/Jhdm1b results in a block at G₁-to-S phase transition. Flow cytometric analysis of cell cycle status shows that Kdm2b/Jhdm1b KD (J1bKD-HMG) results in a higher percentage of cells in the G₁ phase than that of control KD (CKD-HMG).

Figure 4

Kdm2b/Jhdm1 is required for Hoxa9-Meis1-induced AML development in vivo. (A) Flow chart of experimental procedure for BM transplantation assays. To examine the role of Kdm2b/Jhdm1b in Hoxa9-Meis1-induced AML development in vivo, c-kit⁺ progenitors were isolated from E14.5 fetal liver (FL) of Ly5.2 C57BL/6 embryos. After lentiviral (LV) transduction with vectors expressing various combinations of proteins and shRNAs, the genetically modified cells were mixed with normal irradiation protector cells and were transplanted into lethally irradiated Ly5.1 C57BL/6 mice. (B) FACS analysis of the accumulation kinetics of the genetically modified donor cells (GFP, CKD-HMG, and J1bKD-HMG) in the peripheral blood of recipients. The results indicate that Kdm2b/Jhdm1b KD cells (J1bKD-HMG) failed to repopulate in the Mac-1⁺ myeloid lineage, whereas the control KD cells (CKD-HMG) can. (C) The percentage of genetically modified cells, marked by GFP, contribute to the myeloid lineage (Gr-1⁺ and Mac-1⁺) of peripheral blood at various times after transplantation. The results show that Kdm2b/Jhdm1b KD inhibits repopulating by Hoxa9-Meis1-transduced cells in recipient mice. The percentage of GFP⁺ cells in a particular lineage is calculated by dividing GFP and lineage marker double-positive cells with the total lineage marker-positive cells. All error bars represent SD (n = 10). (D) Splenomegaly of mice that received a transplant with lentiviral-transduced BM cells expressing CKD-HMG. Shown is a representative picture of spleens harvested from mice 6 weeks after transplantation of BM cells transduced with lentiviral vectors expressing GFP, J1bKD-HMG, or CKD-HMG. Bar size represents 1.0 cm. (E) H&E staining shows leukemic infiltration of multiple organs (spleen, kidney, and lung) in mice that received a transplant with Hoxa9-Meis1-induced leukemic cells (CKD-HMG), whereas there is no obvious leukemic cell infiltration in mice that received a transplant with BM cells transduced with lentiviral vectors expressing GFP or J1bKD-HMG cells. (F) May-Grünwald/Giemsa staining shows typical leukemic cells in the peripheral blood of mice that received a transplant with CKD-HMG cells. Bar size represents 50 μm. (G) Survival curve shows mice that received a transplant with GFP and J1bKD-HMG cells survived ≥ 70 days after transplantation, whereas the mice that received a transplant with CKD-HMG cells all died within 70 days after transplantation.

Figure 5

The enzymatic activity of KDM2b/JHDM1b is required for the self-renewal of leukemic stem cells. (A) Flow chart of experimental procedure for analyzing the role of Kdm2b/Jhdm1b in LSC self-renewal. Leukemic cells were isolated from primary AML mice and selected for LSCs through replating on methylcellulose. Cells derived from the colonies were transduced with various lentiviral (LV) vectors, followed by methylcellulose colony formation assay in vitro and secondary transplantation assays in vivo. (B) Photographs of the methylcellulose colony formation assay plates show that Kdm2b/Jhdm1b KD results in a decease of both size and number of colonies. This phenotype can be rescued by wild-type but not a catalytic mutant Kdm2b/Jhdm1b or LacZ. (C) Quantification of the colony numbers derived from the methylcellulose colony replating assays. (D) Survival curve shows prolonged survival of mice that received a transplant of primary Hoxa9-Meis1 leukemia cells with KD of Jhdm1b. This phenotype can be reversed by wild-type but not catalytic mutant Kdm2b/Jhdm1b or LacZ. CFC indicates colony-forming cell; Wt, wild-type; and Mut, mutant.

Figure 6

KDM2b/JHDM1b directly regulates p15^Ink4b expression in leukemic cells. (A) RT-qPCR analysis shows that p15^Ink4b is significantly up-regulated in response to Kdm2b/Jhdm1b KD. Relative mRNA levels are measured by RT-qPCR and normalized to Gapdh level. (B) Schematic representation of the p15^Ink4b locus in mouse, indicating the genomic structure (exons are represented by black boxes), as well as the location of the 3 amplicons analyzed by ChIP assays. (C) ChIP experiments with chromatin prepared from leukemic cells expressing LacZ-Flag, wild-type KDM2b/JHDM1b-Flag, and mutant KDM2b/JHDM1b-Flag were carried out with the use of anti-Flag antibody. KDM2b/JHDM1b-Flag binding was assayed by qPCR at the 3 genomic regions depicted in panel B. (D) ChIP experiments with chromatin prepared from leukemic cells with control KD, Kdm2b/Jhdm1b KD, Kdm2b/Jhdm1b KD reconstituted with wild-type Kdm2b/Jhdm1b or mutant Kdm2b/Jhdm1b were carried out with the use of the anti-H3K36me2 antibody. H3K36me2 level was assayed by qPCR at the 3 genomic regions depicted in panel B.

Figure 7

p15^Ink4b is a critical target mediating the function of KDM2b/JHDM1b in Hoxa9-Meis1-induced AML. (A) Loss of p15^Ink4b abrogates the effect of Kdm2b/Jhdm1b KD on Hoxa9-Meis1-induced leukemic transformation. Serial methylcellulose replating assay shows that similar numbers of colonies were obtained from p15^Ink4b-null HPCs transduced with Hoxa9-Meis1 regardless of whether Kdm2b/Jhdm1b was knocked down. (B) Splenomegaly is observed in mice that received a transplant with both control KD (CKD/HMG) and Kdm2b/Jhdm1b KD (J1bKD-HMG) Hoxa9-Mesi1-transduced p15^Ink4b-null hematopoietic progenitors. (C) Flow cytometric analysis shows that both J1bKD-HMG and CKD/HMG transduced p15^Ink4b-null hematopoietic progenitor donor cells can repopulate and dominate peripheral blood of recipient mice 4 weeks after transplantation. (D) Survival curve shows that mice that received a transplant with p15^Ink4b-null HPCs transduced with either CKD-HMG or J1bKD-HMG die within 90 days after transplantation. (E) Proposed model for epigenetic regulation of the Ink4b locus in LSCs. In LSCs, p15^Ink4b is suppressed by multiple epigenetic modifiers, including KDM2b/JHDM1b and Polycomb group proteins. In this model, Kdm2b/Jhdm1b is up-regulated by oncogenic stimuli and maintains at a high level in LSCs. Demethylation of H3K36 by KDM2b/JHDM1b and concomitant H2A ubiquitylation by the associated Polycomb group of proteins results in silencing of p15^Ink4b. Conversely, depletion of Kdm2b/Jhdm1b causes an increase in the H3K36me2 level concomitant with the loss of Polycomb group proteins, leading to de-repression of p15^Ink4b, resulting in defects in leukemic cell proliferation and LSC self-renewal.