Ikaros inhibits megakaryopoiesis through functional interaction with GATA-1 and NOTCH signaling

Abstract

The transcription factor Ikaros regulates the development of hematopoietic cells. Ikaros-deficient animals fail to develop B cells and display a T-cell malignancy, which is correlated with altered Notch signaling. Recently, loss of Ikaros was associated with progression of myeloproliferative neoplasms to acute myeloid leukemia and increasing evidence shows that Ikaros is also critical for the regulation of myeloid development. Previous studies showed that Ikaros-deficient mice have increased megakaryopoiesis, but the molecular mechanism of this phenomenon remains unknown. Here, we show that Ikaros overexpression decreases NOTCH-induced megakaryocytic specification, and represses expression of several megakaryocytic genes including GATA-1 to block differentiation and terminal maturation. We also demonstrate that Ikaros expression is differentially regulated by GATA-2 and GATA-1 during megakaryocytic differentiation and reveal that the combined loss of Ikzf1 and Gata1 leads to synthetic lethality in vivo associated with prominent defects in erythroid cells and an expansion of megakaryocyte progenitors. Taken together, our observations demonstrate an important functional interplay between Ikaros, GATA factors, and the NOTCH signaling pathway in specification and homeostasis of the megakaryocyte lineage.

Figure 1

Ikaros partly inhibits Notch-mediated megakaryocyte differentiation. LSK and CMP cells purified by flow cytometry from wild-type C57BL/6 mice were transduced with an empty retroviral vector (MIE) or encoding full-length Ikaros encoding (IK1) and plated directly on OP9 and OP9-DL1 stroma for 5 and 7 days, respectively. A representative flow cytometry analysis is shown and lower panel bar graphs depict the means ± SD for 3 independent experiments. *P = .04; #P = .007.

Figure 2

Ikaros inhibits growth of AMKL cells transformed with OTT-MAL. (A) The viability of GFP-sorted 6133 cells expressing GFP alone (MIE), full-length Ikaros (IK1), or the IK7 dominant-negative isoform was assessed over time by trypan blue exclusion. Means ± SD of triplicate experiments are shown. (B) c-Kit, CD41, and CD42 surface marker expression 48 hours after transduction of 6133 cells with MIE, IK1, or IK7. (Left panels) Representative experiment. (Right panels) Bar graph of the means ± SD of the percentages of CD41⁺CD42⁺ cells (n = 3). *P = .0004. (C) Annexin V, 7-AAD staining of 6133 cells was evaluated 24 and 48 hours after transduction by flow cytometry. (Top panel) Representative FACS plot at 48 hours. (Bottom panel) Bar graph depicts means ± SD of triplicate experiments. *P < .05. (D) Cell-cycle analysis of GFP- or IK1-expressing 6133 cells was measured by BrdU incorporation and 7-AAD staining coupled with flow cytometry. *P = .0004. (E) Expression of the Notch target gene Hes1, Nrarp, Kit, Cdkn1a, Dtx2, and Hes6 in 6133 cells expressing GFP or IK1 was measured by qRT-PCR. Means ± SD of 3 independent experiments are shown. *P ≤ .01. 7-AAD, 7-amino actinomycin D; FACS, fluorescence-activated cell sorter.

Figure 3

Ikaros binds and represses critical megakaryocytic genes. (A) Heatmap of the 4325 genes differentially expressed in 6133 cells overexpressing IK1. Selected genes of interest are shown. (B) Venn diagram of the overlap between bound and differentially expressed genes in 6133 cells overexpressing Ikaros. (C) Pie chart of the location of peaks relative to the 309 differentially expressed genes that are bound by Ikaros. Promoters are considered to be 2 kb upstream of the TSS. (D) De novo motif finding using 200 bp of sequence surrounding the peak center of the Ikaros peaks. A database of known transcription factor binding motifs was interrogated to find the closest match for each motif. The score indicates the similarity between the found motif and the motif in the database with 1 being a perfect match. (E) UCSC Genome Browser depiction of Ikaros chromatin occupancy in 6133 cells at selected megakaryocyte genes. (F) Western blot of Ikaros and GATA-1 expression in 6133 cells transduced with IK1, IK7, or GFP alone. HSC70 is included as a loading control. Quantification of band intensities relative to Hsc70 and normalized on protein extracted from cells transduced with the empty retroviral vector Migr1 are indicated (also see supplemental Figure 2B).

Figure 4

Validation of Ikaros targets in primary megakaryocytes. (A) Fold changes in expression of the selected Ikaros target genes in megakaryocytic cultures of bone marrow cells. IK1- or IK7-transduced whole bone marrow cells were GFP sorted at 48 hours and cultured under megakaryocytic conditions (SCF/THPO) for 3 days prior to RNA extraction. Genes bound by IK1 are in bold (defined as having >0, but <30 tags), and genes that are stringently bound (defined as having >30 tags) are bold and underlined. Means ± SD from 4 independent experiments are shown. *P ≤ .05; #P ≤ .01; $P ≤ .001. (B) Fold changes in expression of selected megakaryocytic genes in CD41-positive sorted cells isolated from 8- to 9-week-old wild-type, Ikaros^+/−, and Ikaros-null mice. Means ± SD from 3 independent experiments are shown. *P ≤ .02 relative to Ik^+/−.

Figure 5

Ikaros expression is regulated by the GATA switch during megakaryopoiesis. (A) Lineage-negative wild-type bone marrow cells were grown in SCF/THPO media for 4 days and RNA extraction was performed each day. Expression of Ikaros, Gata2, and Gata1 were assessed by qRT-PCR analysis. Data were normalized to day 1. Means ± SD are shown from 4 independent experiments. *P < .05; #P ≤ .01. (B) G1ME cells were transduced with retroviral vectors Migr1 or Migr1-Gata1, GFP sorted, and cultured for 2 days prior to RNA extraction. Data are shown as means ± SD from 3 independent experiments. *P ≤ .005. (C) Genome browser picture of Ikzf1 locus with the indicated GATA-2 (dark blue), GATA-1 (red), and switch (green) binding sites. Data were extracted from GATA-2 and GATA-1 ChIP-Seq experiments. (D) qRT-PCR performed on RNA extracted from day 2 GFP-sorted G1ME transduced with banshee, banshee-shGata2, Migr1, and Migr1-Gata1 retroviral vectors. Means ± SD of triplicate experiments are shown. *P ≤ .01. (E) Western blots performed with nuclear extracts from day 2 GFP-sorted G1ME cells transduced with banshee, banshee-shGata2, Migr1, and Migr1-Gata1 retroviral vectors. Quantification of band intensities relative to Hsc70 and normalized to empty vector Banshee or Migr are indicated.

Figure 6

Functional interaction between Ikaros and Gata1 during fetal hematopoiesis. (A) GATA-2– and GATA-1–dependent Ikaros expression in 13.5-day G1KD fetal liver cells. qRT-PCR assays were performed on GFP-sorted 13.5-day G1KD FL cells transduced with banshee (empty vector), banshee-shGata2, Migr1, and Migr1-Gata1 retroviral vectors. Means ± SD from 5 to 6 independent experiments are shown. *P = .004; #P = .0007. (B) Images of G1KD-Ikaros-null and G1KD-Ik^+/− fetuses at 3 different time points. (C) Histogram plots showing the cell numbers of E13.5 G1KD-Ikaros fetal livers before and after RCL. Data are shown as means ± SD (n = 5-6 fetal livers per group); *P < .001 relative to G1KD-Ikaros^+/+. #P < .001 relative to G1KD-Ikaros^+/−. (D) Bar graphs comparing the percentages of the CD71⁺/Ter119⁻ and CD71⁺/Ter119⁺ positive cells of Ik^+/+, Ik^+/−, and Ik^−/− fetal liver cells in a Gata1 Wt or G1KD background. Data are shown as means ± SD from 6 independent experiments; *P < .001 relative to G1KD-Ik^+/+. #P < .001 relative to G1KD-Ik^+/−. (E) Representative FACS plots of megakaryocytic subpopulations in E13.5 G1KD-Ik^+/− and G1KD-Ik^−/− fetal livers. (F) Histogram plots showing the increased percentage of the c-Kit⁺/CD41⁺ positive cells in the G1KD-Ik^−/− E13.5 fetal livers. Data are shown as means ± SD from 7 independent experiments. *P < .001 relative to G1KD-Ik^+/+; #P = .001 relative to G1KD-Ik^+/−. RCL, red cell lysis; Wt, wild type.

Figure 7

Regulation of megakaryopoiesis by Ikaros. We propose that Ikaros controls megakaryocytic development through 3 different mechanisms: Ikaros inhibits the NOTCH-induced megakaryocytic specification from hematopoietic progenitors, interferes with GATA-1 expression and function to inhibit differentiation and terminal maturation, and directly represses expression of several megakaryocytic genes.