RepID represses megakaryocytic differentiation by recruiting CRL4A-JARID1A at DAB2 promoter

Abstract

Background: Megakaryocytes (MKs) are platelet precursors, which arise from hematopoietic stem cells (HSCs). While MK lineage commitment and differentiation are accompanied by changes in gene expression, many factors that modulate megakaryopoiesis remain to be uncovered. Replication initiation determinant protein (RepID) which has multiple histone-code reader including bromodomain, cryptic Tudor domain and WD40 domains and Cullin 4-RING E3 ubiquitin ligase complex (CRL4) recruited to chromatin mediated by RepID have potential roles in gene expression changes via epigenetic regulations. We aimed to investigate whether RepID-CRL4 participates in transcriptional changes required for MK differentiation.

Methods: The PCR array was performed using cDNAs derived from RepID-proficient or RepID-deficient K562 erythroleukemia cell lines. Correlation between RepID and DAB2 expression was examined in the Cancer Cell Line Encyclopedia (CCLE) through the CellMinerCDB portal. The acceleration of MK differentiation in RepID-deficient K562 cells was determined by estimating cell sizes as well as counting multinucleated cells known as MK phenotypes, and by qRT-PCR analysis to validate transcripts of MK markers using phorbol 12-myristate 13-acetate (PMA)-mediated MK differentiation condition. Interaction between CRL4 and histone methylation modifying enzymes were investigated using BioGRID database, immunoprecipitation and proximity ligation assay. Alterations of expression and chromatin binding affinities of RepID, CRL4 and histone methylation modifying enzymes were investigated using subcellular fractionation followed by immunoblotting. RepID-CRL4-JARID1A-based epigenetic changes on DAB2 promoter were analyzed by chromatin-immunoprecipitation and qPCR analysis.

Results: RepID-deficient K562 cells highly expressing MK markers showed accelerated MKs differentiation exhibiting increases in cell size, lobulated nuclei together with reaching maximum levels of MK marker expression earlier than RepID-proficient K562 cells. Recovery of WD40 domain-containing RepID constructs in RepID-deficient background repressed DAB2 expression. CRL4A formed complex with histone H3K4 demethylase JARID1A in soluble nucleus and loaded to the DAB2 promoter in a RepID-dependent manner during proliferation condition. RepID, CRL4A, and JARID1A were dissociated from the chromatin during MK differentiation, leading to euchromatinization of the DAB2 promoter.

Conclusion: This study uncovered a role for the RepID-CRL4A-JARID1A pathway in the regulation of gene expression for MK differentiation, which can form the basis for the new therapeutic approaches to induce platelet production. Video Abstract.

Keywords: CRL4A; JARID1A; Megakaryocytic differentiation; RepID.

Fig. 1

RepID negatively correlates with the DAB2 expression. A Profiling of transcription levels of genes involved in Wnt signaling pathway in wild-type (WT) or RepID-depleted (KO) K562 cancer cell lines using PCR array. B, C DAB2 transcription is negatively correlated with the RepID. Expression of DAB2 transcripts was compared with RepID transcripts in all cancer dataset based on Cancer Cell Line Encyclopedia (CCLE) database using CellMinerCDB (B). Heatmap of transcription correlation between RepID and DAB2 from CCLE dataset (C). D DAB2 transcription is increased in a RepID-depleted K562 cell lines. Transcription level of DAB2 was examined by qRT-PCR in RepID WT or RepID-depleted condition. Average of transcription level relative to GAPDH was indicated, and error bars represent the standard deviation (***P-value < 0.001). E DAB2 protein is highly expressed in RepID-deficient K562 cell lines. Immunoblot analysis with antibodies against DAB2 was performed and Histone H3 was used as loading control. Relative intensity of DAB2 protein was indicated as fold changes of RepID KO to RepID WT cells

Fig. 2

RepID depletion accelerates megakaryocytic differentiation. A-D Subpopulations of RepID-deficient K562 cells show megakaryocytic phenotype. Representative images of Wright-Giemsa stained RepID WT and KO K562 cells (A). scale bar: 50 µm. Cell size of RepID WT (n = 329) and KO (n = 290) K562 cells was shown in (B). Percentage of multinucleated cells (WT: n = 2367, KO: n = 3803) (C). Transcription levels of megakaryocyte marker including CD41, CD61 and DAB2 in K562 RepID WT and KO cells (D). E–G Expression level of DAB2 protein is highly induced in RepID-deficient K562 cells by PMA treatment. Immunoblot analysis with antibodies against DAB2 was performed using PMA-treated K562 RepID WT and KO cells, and histone H3 was used as loading control (E). The numbers above the bar graph represent the fold increase of DAB2 protein level at the indicated time after PMA treatment compared to 24 h before, and the numbers in the bar graph represent the DAB2 protein expression in RepID KO cells versus RepID WT cells (F). Immunofluorescence analysis of DAB2 expression in CTL or PMA-treated K562 RepID WT or KO cells. White arrows indicate DAB2-positive cells showing multinucleated morphologies (G). scale bar: 50 µm. H, I RepID-deficient K562 cells are accelerated to the megakaryocytic differentiation by PMA treatment. Representative images of Wright-Giemsa stained in PMA-treated K562 RepID WT and KO cells (H). scale bar: 20 µm. Transcripts of megakaryocyte marker (CD41, CD61 and DAB2) or multinucleated ratios in PMA-treated K562 RepID WT and KO cells were examined by qRT-PCR or cell counting, respectively (I). All error bars in this figure represent the standard deviation (*P-value < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, n. s.; not significant)

Fig. 3

RepID-CUL4A dissociates from the chromatin during megakaryocytic differentiation. A-C RepID WD40 domain is required for inhibition of DAB2 expression. Construction of RepID full-length (FL) and lacking WD40 fragments (ΔWD40) (A). Total protein was extracted from RepID-deficient K562 cells transfected with RepID fragments and K562 RepID WT cells as indicated, and immunoblot analysis was performed using DAB2 antibodies. Alpha-tubulin was used as a loading control (B). Quantification of DAB2 levels, normalized to its levels in RepID WT cells. All error bars represent the standard deviation from 3 independent experiments (*P-value < 0.05, ****P < 0.0001, n. s.; not significant) (C). D, E RepID WD40 domain is required for inhibition of megakaryocytic differentiation. Indicated cells were incubated with PMA (1 day) and polyploidy was measured by flow cytometry analysis (D). Percentages of polyploidy (> 4N) are indicated (dark blue) (E). F Stability of DAB2 protein is not altered regardless of RepID expression. K562 RepID WT or KO cells incubated with or without the protein synthesis inhibitor cycloheximide or proteasome inhibitor MG132. Cell lysates were subjected to the immunoblot analysis and relative densitometry was indicated. G Differential changes of transcription level for selected CRL components (CUL1, CUL2, CUL3, CUL4A, CUL4B, CUL5, CUL7, DDB1, RepID) in PMA-treated K562 cells (20 nM, 2 days) in published microarray dataset (GSE57734), or in published RNA-Sequencing database from MKs and normal HSCs in both species (human and mouse) (https://repository.uel.ac.uk/). H RepID and CUL4A dissociates from chromatin during megakaryocytic differentiation. K562 RepID WT or KO cells incubated with PMA were collected at the indicated times, and chromatin-bound fractions were extracted and immunoblot analysis was performed using anti-RepID, anti-CUL4A, anti-CUL4B and anti-DDB1 antibodies. Alpha-tubulin and histone H3 was used for loading as well as fractionation control. Quantification of RepID, CUL4A, CUL4B and DDB1 levels in whole cell lysates and chromatin-bound fractions after normalizing alpha-tubulin or histone H3, respectively. Quantification was presented under the blots

Fig. 4

DAB2 promoter forms euchromatin by dissociation of CRL4A-JARID1A complex during megakaryocytic differentiation. A Chromatin configuration of DAB2 promoter in RepID-deficient K562 cells. ChIP-qPCR was performed using antibodies against H3K4me3 and H3K9me3 in K562 RepID WT or KO cells to examine epigenetic configuration of the DAB2 promoter region. Data are represented as the fold enrichment compared to the IgG control after normalizing with input value. B Correlation between CRL4 components (RepID, CUL4A, CUL4B and DDB1) and histone H3K4 modifiers including methyltransferase and demethylase in all cancer datasets based on CCLE database using CellMinerCDB. Pearson’s correlation R value was presented as heatmap (left panel). Published interactions between CRL4 components and H3K4 modifiers were analyzed using public data from BioGRID database (right panel). C, D JARID1A interacts with CUL4A. A pull-down assay using a CUL4A antibody was performed using whole cell extracts from K562 RepID WT and KO cells and co-precipitated JARID1A, DDB1 and RepID were analyzed by immunoblotting (C). Immunoprecipitation using JARID1A antibodies to pull down CUL4A (D). E CUL4A dissociates from JARID1A during megakaryocytic differentiation. Whole cell extracts from K562 RepID WT and KO cells induced to megakaryocytic differentiation using PMA for 48 h were immunoprecipitated using CUL4A antibodies, and co-precipitated JARID1A was analyzed by immunoblotting and densitometry. F Chromatin binding of JARID1A is highly decreased during megakaryocytic differentiation. The whole cell extracts and chromatin-bound fractions from K562 RepID WT or KO cells incubated with PMA as the indicated times were analyzed by immunoblot analysis using anti-JARID1A, anti-ASH2L, anti-RBBP5 and anti-WDR5 antibodies. Alpha-tubulin and histone H3 was used as the control for loading and fractionation. Quantification of JARID1A levels in whole cell lysates and chromatin-bound fractions after normalizing alpha-tubulin or histone H3, respectively. Quantification was presented under the blots. G Proximity ligation assay for protein interaction between CUL4A and JARID1A in with/without PMA treatment in pre- or no pre-extraction condition. Each red spot represents a single protein–protein interaction and DNA was stained with DAPI. scale bar: 20 µm. Red signal numbers in each cell were quantified by analyzing 315–747 cells randomly, and average numbers for positive red signals per cells are indicated. H, I JARID1A-CUL4A complex is recruited by RepID and dissociates from DAB2 promoter during megakaryopoiesis. ChIP-qPCR of DAB2 promoter region was performed using antibodies against RepID, JARID1A and CUL4A in K562 RepID WT or KO cells treated with PMA to induce megakaryocytic differentiation (H). Re-ChIP-qPCR analysis using RepID antibodies as a first precipitation, followed by second ChIP using rabbit IgG, CUL4A, mouse IgG or JARID1A antibodies (I). All data are presented as the fold enrichment compared to the IgG control after normalizing with input value

Fig. 5

Schematic model. In RepID WT cells, CRL4A forms complex with subset of JARID1A in nucleus and CRL4A-JARID1A complex is recruited to DAB2 promoter mediated by interaction between RepID WD40 domain and CRL4A. JARID1A loading on DAB2 promoter maintains its epigenetic configuration as heterochromatin status, leading to transcription repression. During megakaryocytic differentiation, CRL4A-JARID1A dissociates from the DAB2 promoter due to the reduction of chromatin binding affinity of RepID, followed by dissociation between CRL4A and JARID1A as well as between RepID and CRL4A, that leads to the increase of DAB2 transcription. In RepID depleted (RepID KO) cells, CRL4A also forms complex with subset of JARID1A in nucleus, but recruitment of this complex on DAB2 promoter is compromised due to absence of RepID. Thus, in RepID KO cells, DAB2 transcription is relatively higher than RepID WT cells, and megakaryocytic differentiation is more accelerated in response to PMA treatment