Lineage-specific activators affect beta-globin locus chromatin in multipotent hematopoietic progenitors

Abstract

During development, the regulated expression of tissue-specific genes can be preceded by their potentiation, that is, by chromatin activation in progenitor cells. For example, the human beta-like globin genes are potentiated in a gene- and developmental-specific manner in hematopoietic progenitors. Developmental regulation of human beta-gene expression in erythroid cells is mostly determined by transcriptional activators; however, it is not clear how gene-specific potentiation is set in hematopoietic progenitors. Using human and transgenic multipotent hematopoietic progenitors, we demonstrate that human beta-globin locus activation is characterized by TBP, NF-E2, CBP and BRG1 recruitment at both the Locus Control Region and human beta-gene promoter. Our results further indicate that in hematopoietic progenitors, EKLF influences chromatin organization at the human beta-globin locus and is instrumental for human beta-gene potentiation. Thus, we show that lineage-specific transcriptional activators expressed at basal levels in progenitor cells can participate in gene potentiation.

Figure 1

Expression of marker genes in ln2 HPC and ln2 EKLF^−/− HPC. (A) Semiquantitative RT–PCR performed on equal amounts of RNA purified from ln2 HPC or ln2 EKLF^−/− HPC. PCR samples were resolved onto agarose gel. γ-Globin: fetal human (γ) and mouse embryonic (βH1) transcripts; β-globin: adult human (β and δ) and mouse (βmin and βmaj) globin transcripts; NPM: ubiquitously expressed nucleophosmin transcript, used as internal control; Neg: negative control; Pos: positive control; (B) Western blot analysis of ln2 HPC, ln2 EKLF^−/− HPC and ln2 EryC; 4 and 8 μg of whole-cell protein extract were loaded in each lane of a 10% SDS–PAGE. Anti-GATA-1 and -p45 antibodies were purchased from SantaCruz; anti-actin antibodies were purchased from LabVision; anti-EKLF serum is a generous gift of S Philipsen. Protein levels in ln2 HPC versus ln2 EKLF^−/− HPC or ln2 EryC versus ln2 HPC were calculated using actin as internal control and they are shown on the right side of each panel together with their standard error of means; (C–E) representative examples of quantitative real-time RT–PCR; the relative level of GATA-1 or p45 gene expression in ln2 EKLF^−/− HPC versus ln2 EKLF HPC were calculated according to Pfaffl (2001), using mouse actin as internal control and expressed as KO/WT ratio; x axis: cycle number; y axis: derivative of SYBR Green fluorescence. Blue dots: ln2 HPC; green triangles: ln2 EKLF^−/− HPC.

Figure 2

Histone H3 acetylation, GATA-1 and p45 recruitment at the huβ-globin locus in HPC. (A) A map of the huβ-globin locus; genes are shown as black boxes and the location of βLCR HSs is indicated by arrows. Amplified regions used for ChIP analyses are indicated by grey boxes. (B–G) ChIP assays: immunoprecipitated and unbound (input) chromatin samples were subjected to duplex radioactive PCR with one primer set specific for huβ-globin regions and another primer set specific for mouse zfp37 (ZFP, zinc-finger protein 37) or human pax6 (pax6, paired box protein 6) regulatory regions (‘Ctl', indicated by arrows), two genes that are not expressed in hematopoietic cells. All PCR reactions were performed in parallel under conditions of linear amplification. Products were quantified by PhosphorImager. The level of enrichment of globin regions relative to the control and input samples is represented by bars, with their corresponding standard deviations. A value of 1 indicates no enrichment. Mouse THP/ZFP or human THP/pax6 controls are included (THP, kidney-specific Tamm-Horsfall gene) to confirm that no enrichment is detected at regulatory regions of non-hematopoietic genes. To obviate for weak signals, THP/ZFP PCR reactions were run longer but always in conditions of linear amplification. Representative gel images are shown below each panel; AcH3: anti-acetylated histone H3 antibodies; dark blue bars: ln2 HPC; dark blue hatched bars: human CD34⁺ cells; light blue bars: ln2 EKLF^−/− HPC.

Figure 3

CBP, BRG1, TBP and Pol II recruitment at the huβ-globin locus in HPC. Ln2 HPC, human CD34⁺ cells, or ln2 EKLF^−/− HPC were subjected to ChIP analyses with anti-CPB (A, B), -BRG1 (C, D), -TBP (E, F), or -Pol II (G, H) antibodies. Analysis and quantification of immunoprecipitated samples were performed as described in Figure 2. Dark blue bars: ln2 HPC; dark blue hatched bars: human CD34⁺ cells; light blue bars: ln2 EKLF^−/− HPC.

Figure 4

Quantitative real-time RT–PCR of ln2 HPC and ln2 EKLF^−/− HPC. Representative examples of huβ- (A) and huγ-gene (B) expression were evaluated by real-time RT–PCR. The relative levels of globin gene expression in ln2 EKLF^−/− HPC versus ln2 HPC were calculated according to Pfaffl (2001), using mouse actin as internal control (as described in Figure 1) and expressed as KO/WT ratios. Blue dots: ln2 HPC; green triangles: ln2 EKLF^−/− HPC.

Figure 5

Chromatin organization at the huβ-globin locus in EryC. (A–I) ChIP analyses of ln2 EryC (dark orange bars) or ln2 EKLF^−/− EryC (light orange bars). Analysis and quantification of immunoprecipitated samples, as well as antibodies used for ChIP assays are as described in Figures 2 and 3; MeK4: anti-lysine 4 dimethylated histone H3 antibodies; PCTD: anti-phospho-Ser2 Pol II CTD antibodies.

Figure 6

Model of huβ-globin gene potentiation in ln2 HPC. Schematic representation of huβ-gene potentiation in multipotent ln2 HPC (see Discussion for more details). Nucleosomes are represented by beads and transcriptional activators are indicated as colored objects.