E2A Antagonizes PU.1 Activity through Inhibition of DNA Binding

Abstract

Antagonistic interactions between transcription factors contribute to cell fate decisions made by multipotent hematopoietic progenitor cells. Concentration of the transcription factor PU.1 affects myeloid/lymphoid development with high levels of PU.1 directing myeloid cell fate acquisition at the expense of B cell differentiation. High levels of PU.1 may be required for myelopoiesis in order to overcome inhibition of its activity by transcription factors that promote B cell development. The B cell transcription factors, E2A and EBF, are necessary for commitment of multipotential progenitors and lymphoid primed multipotential progenitors to lymphocytes. In this report we hypothesized that factors required for early B cell commitment would bind to PU.1 and antagonize its ability to induce myeloid differentiation. We investigated whether E2A and/or EBF associate with PU.1. We observed that the E2A component, E47, but not EBF, directly binds to PU.1. Additionally E47 represses PU.1-dependent transactivation of the MCSFR promoter through antagonizing PU.1's ability to bind to DNA. Exogenous E47 expression in hematopoietic cells inhibits myeloid differentiation. Our data suggest that E2A antagonism of PU.1 activity contributes to its ability to commit multipotential hematopoietic progenitors to the lymphoid lineages.

Figure 1

E47 but not EBF interacts with PU.1. (a) The transcription factors ARNT, PAX-5, EBF, and E47 were in vitro translated (IVT) and labeled with ³⁵S-methionine. The labeled proteins were incubated with either GST or GST-PU.1 bacterially produced protein. 10% of IVT proteins and proteins bound to GST and GST-PU.1 were separated by SDS-PAGE and visualized by autoradiography. ARNT is a basic-helix-loop-helix transcription factor, which was used as a negative control (not expected to bind to PU.1). Pax-5 was used as a positive control. (b) E47 and PU.1 are coexpressed in the human hematopoietic cell line K562. Nuclear extracts were prepared from K562 cells and treated with anti-PU.1 antibody or anti-GM-CSFRα prebound to protein A agarose. Isolated complexes were separated on an SDS-PAGE gel and blotted. The presence of E47 in complexes was assayed using an anti-E47 antibody.

Figure 2

Identification of domains required for mediating the interaction between PU.1 and E47. (a) In vitro translated ³⁵S-methionine labeled full-length human E47 (FL, aa 1–654), E47 C-terminal truncation 1 (CT1, aa 1–492), and E47 C-terminal truncation 2 (CT2, aa 1–369) were incubated with bacterially produced GST-PU.1 fusion protein. 10% input proteins and proteins bound to GST-PU.1 were separated by SDS-PAGE and visualized by autoradiography. (b) ³⁵S-methionine labeled in vitro translated PU.1 and E47 were incubated with GST-E47bHLH protein (E47, aa 521–623). In vitro translated E47 was used as a positive control since it should dimerize with the E47 bHLH domain. (c) ³⁵S-methionine labeled in vitro translated PU.1 deletion mutants were incubated with GST-E47bHLH protein.

Figure 3

E47 inhibits promoter activity induced by PU.1. (a) Transient transfection of U937 cells with a MCSFR promoter and GCSFR promoter luciferase construct in the presence or absence of MigR1-E47 plasmid. (b) MCSFR promoter activity induced in 293 T cells with cotransfection of MigR1-PU.1 in the presence or absence of MigR1-E47 plasmid. MigR1-E47 by itself had no effect on MCSFR promoter activity. (c) (μE3)₄-Luciferase activity induced by cotransfection with MigR1-E47 in the presence and absence of MigR1-PU.1. For (a) luciferase activity is reported as fold repression compared to activity in the absence of MigR1-E47. For (b) and (c) luciferase activity is reported as fold-induction above the activity seen in 293 T cells transfected with only the reporter plasmid and MigR1. Luciferase values are the mean ± standard deviation of three independent transfections.

Figure 4

E47 inhibits myeloid differentiation of hematopoietic cells. (a) PUER cells were superinfected with MigR1 or MigR1-E47 virus and pure GFP expressing clones generated. Cell lines were either undifferentiated (no OHT) or differentiated (100 nM OHT) for 8 d. Immunohistochemistry of adherent PUER cell lines induced to differentiate with 100 nM OHT for 8 d. Cells were stained either with the macrophage differentiation marker F4/80 or with normal rat serum (NRS). Magnification is 400x. (b) F4/80 surface expression as determined by flow cytometry of independent PUER GFP and PUER E47 lines with no addition of OHT and after 8 d of OHT treatment. (c) Flow cytometry analysis of lineage minus hematopoietic progenitors infected with control MigR1 virus or MigR1-E47 virus. Cells were cultured for 12 d on OP9 stromal cell layer in the presence of 1 ng/mL IL-7 and 5 ng/mL Flt3L. Differentiation was analyzed by staining cells with anti-CD11b (myeloid) and anti-CD19 (B cell). Percentages of total cells positive for CD11b and CD19 expression are shown in the upper right corner of FACs plots.

Figure 5

E47 inhibits PU.1 binding to DNA. (a) Transient transfections of 293 T cells with a GAL4 responsive luciferase reporter construct. Cells were cotransfected with either GAL4 full-length PU.1 expression plasmid or GAL4-PU.1 activation domain in the presence or absence of MigR1-E47 plasmid. Luciferase activity is the mean ± standard deviation of three independent transfections and luciferase activity is reported as fold-induction above the activity seen in 293 T cells transfected with the GAL4 DNA binding domain expression plasmid. (b) Immunoblot of nuclear extracts prepared from 293 T cells transfected with empty expression plasmids (NE1), PU.1 expression plasmid (NE2), and PU.1 and E47 expression plasmids (NE3, NE4). NE4 was prepared from cells transfected with an increased amount of E47 expression plasmid (++). Immunoblots were probed with anti-PU.1 or anti-E47 antibody as indicated. (c) EMSA performed with nuclear extracts from (b) and a ³²P-labeled MCSFR probe. The PU.1 containing complex was identified by performing competitions with unlabeled wildtype and mutant MCSFR probes as indicated. Additionally the PU.1 containing complex could be ablated by inclusion of anti-PU.1 antibody in the binding mix.

Figure 6

E47 blocks PU.1 from binding to DNA in hematopoietic cells. (a) Immunoblot of PUER and E47 protein expression in nuclear extracts prepared from PUER GFP and PUER E47 cell lines which were grown in the absence or presence (+) of 100 nM OHT. (b) Electrophoretic mobility shift assay (EMSA) with a ³²P-labeled DNA probe from the MCSFR promoter. Nuclear extracts were prepared from PUER GFP and PUER E47 cell lines that were grown for 2 d in the absence or presence (+) of 100 nM OHT. SS indicates supershifted complex. (c) Chromatin immunoprecipitation (ChIP) of the MCSFR promoter from PUER GFP and 2 independent PUER E47 cell lines using anti-PU.1 antibody. PUER cell lines were treated for 7 d with OHT. Chromatin also precipitated with normal rabbit serum (NRS) to demonstrate specificity of the anti-PU.1 immunoprecipitations.