EVI2B is a C/EBPα target gene required for granulocytic differentiation and functionality of hematopoietic progenitors

Abstract

Development of hematopoietic populations through the process of differentiation is critical for proper hematopoiesis. The transcription factor CCAAT/enhancer binding protein alpha (C/EBPα) is a master regulator of myeloid differentiation, and the identification of C/EBPα target genes is key to understand this process. Here we identified the Ecotropic Viral Integration Site 2B (EVI2B) gene as a direct target of C/EBPα. We showed that the product of the gene, the transmembrane glycoprotein EVI2B (CD361), is abundantly expressed on the surface of primary hematopoietic cells, the highest levels of expression being reached in mature granulocytes. Using shRNA-mediated downregulation of EVI2B in human and murine cell lines and in primary hematopoietic stem and progenitor cells, we demonstrated impaired myeloid lineage development and altered progenitor functions in EVI2B-silenced cells. We showed that the compromised progenitor functionality in Evi2b-depleted cells can be in part explained by deregulation of cell proliferation and apoptosis. In addition, we generated an Evi2b knockout murine model and demonstrated altered properties of hematopoietic progenitors, as well as impaired G-CSF dependent myeloid colony formation in the knockout cells. Remarkably, we found that EVI2B is significantly downregulated in human acute myeloid leukemia samples characterized by defects in CEBPA. Altogether, our data demonstrate that EVI2B is a downstream target of C/EBPα, which regulates myeloid differentiation and functionality of hematopoietic progenitors.

Figure 1

EVI2B is expressed in myeloid cells and upregulated upon p42 C/EBPα-ER activation. (a) Flow cytometric analysis of EVI2B expression in different populations of human buffy coat samples. The y axis is shown in logarithmic scale and represents median fluorescence intensity of EVI2B signal. The x axis shows different hematopoietic populations. Results represent average measurement of four buffy coats. (b–d) Quantitative RT-PCR and western blot analysis of K562 cells overexpressing (b) p42 C/EBPα-ER, (c) ER or (d) p30 C/EBPα-ER treated with 1 μM β-estradiol (black bars) or EtOH vehicle control (gray bars). The y axes represent relative human EVI2B mRNA expression compared to vehicle control treatment. The x axis indicates hours upon stimulation. RT-PCR results represent the average of two independent experiments, each done in duplicate. Western blots: upper panels show murine EVI2B staining and lower panels α-tubulin loading control. Positions of m.w. standards (kDa) are shown on the left

Figure 2

EVI2B is a direct C/EBPα target gene. (a) Human C/EBPα ChIP-seq data in K562 C/EBPα-ER expressing cells treated with 1 μM β-estradiol (α-E₂) or vehicle control (α-VC), and K562 ER control cells treated with 1 μM β-estradiol (ER-E₂) for 6 h. Depicted are potential C/EBPα binding sites (peak 1 and 2). Black arrow indicates transcriptional start site and direction of transcription. (b) ChIP-PCR was performed with K562 C/EBPα-ER expressing cells treated as described in panel A. Immunoprecipitation was performed using antibodies against ER (α-ER) or IgG control antibodies (α-IgG). Q-PCR was performed using oligos in peak 1, peak 2, or a 3'UTR control region (3'ctrl region). The y axis indicates percentage of enrichment from input. (c) ChIP-seq dataset (GSM722424) on U937 human cells assessing the binding of endogenous CEBPα within the EVI2B promoter (peak 1.1 and 1.2). (d) ChIP-PCR assay in U937 cells using seven primer sets covering the promoter region, including the two peaks shown by the U937 ChIP-seq (peak 1.1 and peak 1.2), the region between the peaks (mid) along with the 5' and 3' proximal (shoulders) and distal (extreme) regions flanking the peaks. Error bars represent the standard deviation across the replicates (n=3). CEBPα ChIP enrichment on peak 1.1 and 1.2 regions is significantly higher compared to the distal control regions (P=0.0006 and 0.0013, respectively). (e) Luciferase reporter assay in K562 cells co-transfected with 100 ng EVI2B luciferase reporter vector (peak 1 and 2) and increasing amounts of human CEBPA expression plasmid. The y axis indicates the relative percentage of luciferase activity and it is presented as fold change compared to the empty vector (not containing CEPBA). Results represent the average of two independent experiments, each done in triplicate. (f) Luciferase reporter assays in K562 cells transfected with WT, single mutants, double mutants or a triple mutant of human EVI2B promoter peak 1 luciferase reporter vector and 100 ng CEBPA expression plasmid. The x axis indicates the relative percentage of luciferase activity. Results represent the average of at least three independent experiments. Right part of the Figure graphically depicts human EVI2B promoter peak 1, and 3 C/EBP consensus binding sites (red dots). Black crosses indicate target mutations in C/EBP consensus binding sites from CAAT to CCCT. Numbers on the right side of the panel represent distance from TSS

Figure 3

Human EVI2B promoter is transactivated by full-length C/EBPα p42 but not C/EBPα p30 and C/EBPα mutants. (a and b) Luciferase reporter assays assessing transactivation of EVI2B peak 1 by different C/EBPα mutant forms. (a) rat C/EBPα p42 and rat C/EBPα p30, (b) human C/EBPα mut 10, human C/EBPα mut 22N, human C/EBPα mut 22C, and human C/EBPα mut 128. The x axes show increasing concentration of CEBPA construct or its mutant forms (0, 50, 100, 200, 400, 800 ng), the y axes indicate relative reporter activity compared to empty vector control. Results represent the average of three independent experiments. (c) Relative expression of EVI2B mRNA in 493 patient samples with non-affected CEBPA (CEBPA WT), 26 CEBPA mutant cases and 10 CEBPA silenced cases

Figure 4

Evi2b knockdown blocks granulocytic differentiation in mouse 32D/G-CSF-R cells. (a and b) Quantitative RT-PCR in 32D/G-CSF-R cells treated with G-CSF (100 ng/ml) for 8 days. The y axes represent relative expression of (a) Evi2b and (b) Cebpa compared to Gapdh. The x axes indicate days of treatment. Results represent the average of two independent experiments. (c) Quantitative RT-PCR in 32D/G-CSF-R cells transduced with distinct shRNA lentiviral constructs 5 days after infection. The y axis indicates fold change of Evi2b mRNA level after transduction with Evi2b silencing shRNA (Sh3, Sh4) compared to non-silencing controls (NSC1, NSC2). (d) Western blot analysis in 32D/G-CSF-R cells after transduction with the distinct shRNA lentiviral constructs. Upper panel shows murine EVI2B staining and lower panel Actin loading control. Positions of m.w. standards (kDa) are shown on the left. (e) Differential counting of mature neutrophils (white bars), immature neutrophils (gray bars), and blasts cells (black bars) in control and Evi2b-silenced 32D/G-CSF-R cells treated with G-CSF (100 ng/ml) for 7 days in May-Grünwald Giemsa stained cytospins. (f) Percentage of mature neutrophils during the course of G-CSF-induced neutrophilic differentiation of 32D/G-CSF-R cells. Samples were cytospun every day during 7 days of treatment. The x axis demonstrates the time of treatment (days), the y axis represents the percentage of neutrophils. (g) Representative pictures of 32D/G-CSF-R cells treated with G-CSF for 7 days, cytospun and stained with May-Grünwald Giemsa. Black arrows indicate blasts. Original magnification, x40 (e–g) Experiments were performed three independent times and at least 100 cells were counted for each sample

Figure 5

Evi2b knockdown impairs murine HSPC granulocytic colony forming ability. (a) Quantitative RT-PCR in different sorted murine BM hematopoietic populations. The y axis represents relative expression of Evi2b compared to Gapdh control. The x axis indicates hematopoietic populations: long-term HSC (LT), short-term HSC (ST), LKS, common myeloid progenitors (CMP), common lymphoid progenitors (CLP), megakaryocyte erythroid progenitors (MEP), granulocyte monocyte progenitors (GMP), granulocytes (Gr), B cells, and T cells. Values are average of 3 independent experiments. (b) Quantitative RT-PCR of sorted LKS cells transduced with controls (NSC#1, NSC#2) and Evi2b-downregulating shRNAs (sh#3, sh#4) 3 days after infection. The y axis represents relative expression of murine Evi2b mRNA compared to Gapdh. (c) Quantification of myeloid colonies formed by LKS cells infected with the shRNA constructs growing in semisolid medium for 14 days. The y axis indicates number of myeloid colonies per plate. (d) Two representative pictures of colonies formed by NSCs or Evi2b-downregulated LKS cells. Original magnification, x4. Colony culture assays were performed at least three times. (e) Number of GFP⁺ Gr1/Mac1⁺ granulocytic cells per plate in semisolid cultures 14 days after seeding LKS cells transduced with the distinct shRNA lentiviral constructs

Figure 6

Evi2b knockdown impairs repopulating ability of HSPC. (a) Representative dot plot showing engraftment of donor CD45.2⁺ cells transduced with control (NSC1 and NSC2) or Evi2b-silenced (Sh3 and Sh4) constructs and support CD45.1 spleen cells (upper panels). GFP⁺ cells within the CD45.2⁺ gate are shown in the lower panels. Numbers indicate percentage of cells. Blood analysis was performed 4 weeks after transplantation. (b) Quantification of contribution of control or Evi2b-silenced LKS cells to hematopoietic system of the recipient animals 4 weeks after transplantation. The y axis indicates the percentage of GFP⁺ cells out of CD45.2⁺ gate. Each group contains at least 10 animals. (c) Number of GFP⁺ granulocytes derived from the donor (CD45.2⁺) cells. The Y axis indicates the number of GFP⁺Gr/Mac1⁺ × 10³ granulocytes per 100.000 Ly5.2⁺ cells. Each group contains at least five animals. The values on y axes of (b) and (c) were normalized to the original percentage of GFP⁺ cells measured 2 days after infection. Transplantation experiments were performed three times. (d) Representative contour plots of cell cycle analysis on murine LKS cells transduced with controls NSC1 and NSC2 or Evi2b-downregulating shRNAs Sh3 and Sh4. The x axis indicates Hoechst 33342 staining and the y axis shows Pyronin Y staining. The x and y axes are displayed in linear scale. Numbers indicate the percentage of cells in G₀ (lower left quadrant), G₁ (upper left quadrant), and G₂/S/M phase (upper right quadrant). Right panel indicates quantification of cell cycle analysis in LKS cells transduced with the distinct shRNAs. Black bars indicate percentage of cells in G₀, gray bars indicate percentage of cells in G₁, and white bars the percentage of cells in G₂/S/M. Results are average of three independent experiments

Figure 7

Bone marrow cells from Evi2b KO mice present impaired repopulating ability in vivo and impaired plating capacities in vitro. (a) Quantitative RT-PCR in Evi2b WT (+/+), heterozygous (+/−) and KO (−/−) BM nucleated cells. The y axis represents relative Evi2b expression compared to Gapdh control. Each group contains 3 mice. (b) EVI2B expression in EVI2B WT, heterozygous and KO mice assessed on protein level using Western blotting. Equal amounts of cells (0.5 × 10⁶cells) were loaded per lane. Antibodies against murine EVI2B were used to evaluate EVI2B expression in Evi2b KO, heterozygous and WT mice (upper panel). Equal loading was verified by staining the membrane with antibody against Actin (lower panel). Positions of m.w. standards (kDa) are shown on the left. (c) Quantification of colonies formed by 10 000 whole BM cells from WT and Evi2b KO mice plated in semisolid medium M3231 supplemented with G-CSF (100 ng/ml). The y axis represents the number of colonies per plate. Counting was performed 10 days after plating. Results represent the average of 6 animals. (d) Quantification of Gr1/Mac1⁺ cells in cultures from (c). The y axis represents the number of Gr1/Mac1⁺ cells per plate × 10³. (e) Re-plating assays of WT (black bars) and Evi2b KO (white bars) BM cells using Methocult M3434. Initially 5000 BM nucleated cells were plated per well and 2000 cells were re-plated. Colonies were quantified under microscope and re-plated after 10–14 days in semisolid medium. The y axis indicates the number of colonies per plate. This is a representative experiment out of 2. Results indicate the average of 3 independent cultures. (f) Flow cytometric analysis of blood of lethally irradiated recipient mice transplanted with 100.000 WT or Evi2b KO BM nucleated cells along with 0.5 × 10⁶ BM cells. The y axis indicates the percentage of donor (CD45.2⁺) cells in the blood of recipients 4 weeks after transplantation. Each group contains at least 11 animals. Results include data from three independent experiments

Figure 8

EVI2B silencing blocks granulocytic differentiation in human cell lines and primary HSPC. (a) Representative histogram plots showing EVI2B (upper) and CD11b (lower) levels in NB4 cells transduced with NSC or EVI2B silencing shRNA (Sh5). The y axes indicate cell counts normalized to mode. Right panels indicate mean fluorescence intensity (MFI) and frequency of positive cells in the infected GFP⁺ and the control GFP⁻ populations. (b) Representative histogram plots showing CD11b and CD15 staining in human CD34⁺ cells infected with NSC or EVI2B silencing shRNA (Sh5 and Sh6) upon 10 days in culture. The y axes are shown in modal display mode. Right panels show MFI and frequency of positive cells for CD11b and CD15