The polycomb group protein L3MBTL1 represses a SMAD5-mediated hematopoietic transcriptional program in human pluripotent stem cells

Abstract

Epigenetic regulation of key transcriptional programs is a critical mechanism that controls hematopoietic development, and, thus, aberrant expression patterns or mutations in epigenetic regulators occur frequently in hematologic malignancies. We demonstrate that the Polycomb protein L3MBTL1, which is monoallelically deleted in 20q- myeloid malignancies, represses the ability of stem cells to drive hematopoietic-specific transcriptional programs by regulating the expression of SMAD5 and impairing its recruitment to target regulatory regions. Indeed, knockdown of L3MBTL1 promotes the development of hematopoiesis and impairs neural cell fate in human pluripotent stem cells. We also found a role for L3MBTL1 in regulating SMAD5 target gene expression in mature hematopoietic cell populations, thereby affecting erythroid differentiation. Taken together, we have identified epigenetic priming of hematopoietic-specific transcriptional networks, which may assist in the development of therapeutic approaches for patients with anemia.

Figure 1

KD of L3MBTL1 Primes the Hematopoietic Potential of iPSCs (A) Endogenous L3MBTL1 expression was assessed in iPSCs at different time points during the mesodermal, hematopoietic, and erythroid differentiation using quantitative real-time PCR. The data represent the mean ± SD of the three independent experiments. (B) Strategy diagram. L3MBTL1 expression is efficiently knocked down as assessed by western blot assay in undifferentiated iPSCs. Tubulin served as the loading control. (C) KD of L3MBTL1 increases the expression of mesodermal-specific transcription factors while it decreases expression of key markers of endodermal and ectodermal lineages, as shown in L3MBTL1-KD undifferentiated iPSCs by qPCR compared to controls. Data indicate the relative expression level of the gene of interest, normalized by GAPDH. The reported fold changes have been calculated by ΔΔCt analysis versus control cells. The data represent the mean ± SD of the three independent experiments. ^∗∗p < 0.01 by Student’s t test. (D) KD of L3MBTL1 increases CFU capacity of iPSC-derived HSCs. Cells from day 10 hEBs were plated in methylcellulose and colonies were scored after 15 days. The data represent the mean ± SD of the three independent experiments. ^∗p < 0.05 by Student’s t test. (E) KD of L3MBTL1 promotes the emergence of early hemogenic precursor cells. EBs were harvested at day 10, prior to CD45 emergence, and analyzed by flow cytometry for CD31 and CD34 expression. Refer to Figure S1 for the characterization of the iPSC line generated from cord blood CD34⁺ cells.

Figure 2

Impaired Development of Neural Progenitors in L3MBTL1-KD Pluripotent Stem Cells (A) KD of L3MBTL1 impairs the ability to form neuronal progenitor cells and neurons. Nestin and Tuj1 expression were evaluated in EBs by IF. DAPI served as control. (Top) The first line shows multiple EBs at lesser magnification (4×); the second line shows a single EB at 10× magnification. (Bottom) The impairment in neural progenitor formation of the L3MBTL1-KD EBs. The graph on the right shows the percentage of EBs containing Tuj1⁺ cells. The data represent the mean ± SD of the three independent experiments. ^∗p < 0.05 by Student’s t test. (B) KD of L3MBTL1 enhances CD34 expression. (Top left) The first line shows few positive CD34⁺ EBs at low magnification (4×) in the controls; the second line shows a single EB at higher magnification (10×). (Bottom left) Increased CD34 staining in the L3MBTL1-KD EBs at 4× and 10× magnification is apparent. (Top right) Increased expression of CD34 marker (21%) in the L3MBTL1-KD cells compared to controls (3%) by FACS is shown. (Bottom right) The graph gives the percentage of EBs containing CD34⁺ cells. The data represent the mean ± SD of the three independent experiments. ^∗p < 0.05 by Student’s t test. See also Figure S2.

Figure 3

L3MBTL1 Transcriptionally Represses SMAD5 (A) GEP of undifferentiated L3MBTL1-KD iPSCs, compared to controls, based on microarray analysis. We utilized independent clones of iPSCs (generated from cord blood CD34+ cells), which we independently infected with lentiviral vectors expressing shRNAs against L3MBTL1. (B) mRNA expression levels of several upregulated and downregulated genes, identified by microarray analysis, was confirmed by qPCR. Data were normalized by GAPDH expression and shown as shRNA versus control. The data represent the mean ± SD of the three independent experiments. (C) GSEA for the combined set of SMAD and hematopoietic stem cells genes compared with the differentially expressed genes in L3MBTL1-KD iPSCs. (D) Expression levels of phospho SMAD1/5, total SMAD5, phospho SMAD2, total SMAD2, and HHEX were evaluated in L3MBTL1-KD and control iPSCs by western blot. Tubulin served as the loading control. (E) iPSCs were crosslinked with 1% formaldehyde and immunoprecipitated with anti-L3MBTL1 or IgG antibody (as a nonspecific control). Plotted values are relative enrichments (y axis) to 10% input and measured for sites in the SMAD5 promoter and actin (x axis). The data represent the mean ± SD of the three independent experiments. ^∗p < 0.05 by Student’s t test. (F) L3MBTL1-KD and control iPSCs were crosslinked with 1% formaldehyde and immunoprecipitated with an anti-H3K27 tri-methyl antibody or an IgG antibody (as a nonspecific control). Plotted values are relative enrichments (y axis) to 10% input and measured for sites in the SMAD5 promoter (x axis). The data represent the mean ± SD of the three independent experiments. ^∗p < 0.05 by Student’s t test. See also Figure S3.

Figure 4

Regulation of SMAD5 by L3MBTL1 Occurs in Erythroid Cells (A) KD of L3MBTL1 induces early expression of the erythroid-specific marker GlyA in iPSC-derived CD34+ cells. Representative flow cytometry analysis of iPSCs after 2 days in Epo-induced culture. (B) Expression levels of EKLF, NF-E2, LMO2, EPOR, and fetal globin gene were assessed in iPSC-derived erythroid progeny by qRT-PCR. Data were normalized by GAPDH expression. The data represent the mean ± SD of the three independent experiments. ^∗∗p < 0.01 by Student’s t test. (C) K562 erythroleukemia cells were retrovirally transduced to express L3MBTL1-HA (blue bars) or empty MIGR1 control (gray bars), crosslinked with 1% formaldehyde, and immunoprecipitated with an anti-SMAD5 antibody or IgG antibody. Primers covering the SMAD-binding motifs (Adelman et al., 2002) across the upstream enhancer, the proximal promoter, and the intronic enhancer of the EKLF were utilized. Data were normalized by 10% input. The data represent the mean ± SD of the three independent experiments. (D) EKLF, NF-E2, GATA-1, LMO2, EPOR, and BCL11A expression levels were evaluated by qRT-PCR in L3MBTL1-KD CB CD34⁺ cells after 7 days of erythroid-supporting culture (100 ng/ml SCF and 6 U/ml EPO) compared to controls. GAPDH served as a housekeeping gene control. The data represent the mean ± SD of the three independent experiments. ^∗∗p < 0.01 by Student’s t test. (E) Fetal globin levels were evaluated in L3MBTL1-KD CB CD34⁺ cells by western blot. Tubulin served as the loading control. (F) Overexpression of L3MBTL1 decreases globin gene mRNA levels in cord blood CD34⁺ cells, as shown by qRT-PCR. Cells were cultured for 3 days in liquid culture that supports erythroid differentiation. GAPDH served as the housekeeping gene. The data represent the mean ± SD of the three independent experiments. ^∗∗p < 0.01 by Student’s t test. (G) Overexpression of L3MBTL1 dramatically decreases protein expression levels of gamma globin in K562 cells, as shown by western blot assay. Tubulin served as the loading control. See also Figure S4.