A positive role for NLI/Ldb1 in long-range beta-globin locus control region function

Abstract

Long-range interactions between distant regulatory elements, such as enhancers, and their target genes underlie the specificity of gene expression in many developmentally regulated gene families. NLI/Ldb1, a widely expressed nuclear factor, is a potential mediator of long-range interactions. Here, we show that NLI/Ldb1 and erythroid-binding partners GATA-1/SCL/LMO2 bind in vivo to the beta-globin locus control region (LCR). The C-terminal LIM interaction domain of NLI is required for formation of the complex on chromatin. Loss of the LIM domain converts NLI into a dominant-negative inhibitor of globin gene expression, and knockdown of NLI by using shRNA results in failure to activate beta-globin expression. Kinetic studies reveal that the NLI/GATA-1/SCL/LMO2 complex is detected at the beta-globin promoter coincident with RNA Pol II recruitment, beta-globin transcription, and chromatin loop formation during erythroid differentiation, providing evidence that NLI facilitates long-range gene activation.

Figure 1. The human β-globin LCR is a natural target of E box/GATA binding proteins

(A) The human β-globin locus is illustrated. Vertical arrows denote hypersensitive sites. The positions of TaqMan probes used for real-time qPCR are indicated below and named on the graph. Primer HS2/E box amplifies the HS2 E box/GATA motif (see Experimental Procedures). E19254 and E22236 are E box/GATA motifs upstream and downstream of the ε-globin gene, respectively. IVR denotes a position between the LCR and the ε-globin gene. Chromatin was prepared from 1% formaldehyde cross-linked K562 cells, digested with MNase and sonicated and then subjected to ChIP using antibodies to NLI, SCL, GATA-1, Lmo2 or control IgG. The results for at least three chromatin preparations are shown ± SEM. (B) A 1.46-kb HS2 fragment containing the HS2 core with 5' and 3' flanking sequences was linked to a complete genomic ε-globin gene on minichromosomes in the wild type (WT) construct. E box mutant and GATA mutant were identical except the E box or tandem GATA sites in the HS2 core were mutated by clustered point mutations. Gray boxes, exons of the gene. The positions of TaqMan primers used for real time PCR are indicated beneath the constructs. ChIP assays were performed using K562 cells stably carrying wild-type, E box mutant, or GATA mutant construct as described above. The results for at least three chromatin preparations are shown ± SEM.

Figure 2. Enhancer activity of the E box/GATA complex on ε-globin expression

(A) RT-qPCR analysis of ε-globin transcription in K562 cells carrying HS2ε minichromosomes with WT HS2 or with HS2 mutated to destroy the E box or GATA sites (see Figure 1B). Graphed are the averages of three RNA preparations ± SEM. Data for the different clones were normalized to the actin signal. (B) Wild-type (WT) or C terminal deleted NLI mutant (1−285) proteins are depicted. K562 cells were stably co-transfected with HS2ε minichromosomes (Kim and Dean, 2003) and HA-tagged-wild-type (WT) or C terminal deleted NLI mutant constructs (1−285). NLS, nuclear localization signal. Nuclear extract from stable cell lines were immunoprecipitated by anti-HA agarose antibody, and input proteins or precipitates were analyzed by western blot analysis with antibodies shown on the left. Ctrl., control plasmid without insert. (C) RNA was isolated from stable cell lines and minichromosomal ε-globin mRNA was detected by RNase protection assay. The mean ε-globin mRNA normalized to actin RNA (± SEM) for several clones of each type is illustrated graphically. (D) ChIP assays were carried out using stable cell lines expressing either HA-tagged WT or truncated (1−285) NLI with anti-HA antibody as described for Figure 1. The results for at least three chromatin preparations are presented (± SEM).

Figure 3. NLI knock-down inhibits MEL cell differentiation

(A) Reduced NLI expression in stable NLI shRNA clones. MEL cells were stably transfected with control or NLI shRNA. Western blot analysis was performed with the indicated antibodies. Actin served as an internal control. #13, #14, and #17 represent individually derived NLI shRNA stable clones. (B) Western blot analysis of NLI expression during differentiation. Stable clones were treated with 1.5 % of DMSO and whole cell extracts were prepared at the indicated times. α-tubulin served as a loading control. Similar results were obtained from independent experiments with other clones. (C) Cell viability of stable clones during differentiation. Cells were cultured with 1.5 % DMSO, and cell viabilities were assessed daily by the trypan blue dye exclusion method (averages of 3 clones ± SEM). (D) Cells were treated with 1.5 % DMSO for 4 days and then hemoglobinization was assayed by staining with benzidine. (E) Stable clones with control or NLI shRNA were treated with or without 1.5% DMSO. Total RNA was isolated at day 4, and ε^y, βh1, and β-major mRNA expression was analyzed by quantitative real-time PCR (averages of 3 clones ± SEM). 18S ribosomal RNA served as a control.

Figure 4. Stabilization of Lmo2 protein through the LIM domain of NLI

(A) Increasing amounts of wild type or mutant NLI (5ug and 10ug) expression vectors were co-transfected in to 293 cells with a fixed amount of Lmo2 construct (5ug). After 48hr, whole cell extracts were prepared. Western blotting was performed with the indicated antibodies. (B) Nuclear extracts prepared from 293T cells transfected with the indicated constructs for 48 hr. The immunoblot was incubated with antibodies shown on the left. (C) 293T cells were transfected with the indicated constructs for 48hr. Whole cell extracts were prepared and western blots were incubated with the indicated antibodies

Figure 5. Enrichment of E box/GATA complex on β-globin promoter region during mouse MEL cell differentiation

(A) MEL cells were treated with 2% DMSO and analyzed over a 4 day time course and globin mRNA was analyzed at each stage of differentiation by quantitative real-time PCR (± SEM). 18S ribosomal RNA served as a control. (B) Proteins were analyzed by western blot with specific antibodies described in the legend to Figure 3. Actin served as loading control. (C) The mouse β-globin locus is diagrammed and the positions of TaqMan probes used for real-time PCR in panels D-G are indicated below and named on the graphs. (D) Chromatin was prepared from MEL cell treated with 2% DMSO over 4 days. ChIP assays were performed as described in the legend for Figure 1 with antibodies against RNA pol II. The color key for each time point is the same as in (A). The results are averages of three chromatin preparations ± SEM. (E-G) The same chromatin preparations used for pol II ChIP were analyzed with antibodies against NLI (E), GATA-1 (F) or Lmo2 (G). The control IgG value is plotted for time 0 and was similar at all time points.

Figure 6. NLI and Lmo2 occupy the LCR and β-globin promoter in E14.5 mouse fetal liver cells

Single cell suspensions of E14.5 fetal liver and brain cells were prepared. Cross-linking and ChIP was then carried out as for MEL cells using antibodies to NLI and Lmo2 or IgG as control. The results are averages of two chromatin preparations ± SEM. (A) NLI and Lmo2 occupancy in E14.5 fetal liver. (B) NLI and Lmo2 occupancy in E14.5 fetal brain.

Figure 7. Spatial communication between the β-globin LCR and gene depends on E box/GATA complex occupancy

(A) The murine β-globin locus and the Hind III fragments used in this study. (B) Relative proximity of HS2 with different globin genes in MEL cells (top) before DMSO (0) and after 4 days of DMSO treatment (n=3, ± SEM). All PCR products were identified as the expected ligation dependent products by sequencing. Quantitation of the data was by real-time qPCR. Signals were normalized to a control template and an interaction in the ERCC locus to adjust for template amount and quality. In the lower graph, proximity of HS2 and globin genes in MEL cells stably expressing either control shRNA or NLI shRNA are compared. The 3C experiment was carried out for three days to assure survival of NLI shRNA cells. (C) The ChIP-loop assay was performed by digesting cells with Hind III after 4 days of DMSO induction (IMEL, upper panel) or with no induction (UMEL, lower panel). Chromatin was immunoprecipitated with antibodies against NLI or control IgG followed by ligation and PCR amplification with 3C primers. Ctrl. represents the PCR control template of the size expected for amplification between the HS2 primer and the β-maj primer or other globin genes where NLI interaction with chromatin was not observed (Figure 5 and data not shown). (D) A model of long range communication between the β-globin LCR and gene facilitated by E box/GATA complex formation through the self-interacting domain of NLI/Ldb1.