Distinct C/EBP functions are required for eosinophil lineage commitment and maturation

Abstract

Hematopoietic differentiation involves the commitment of multipotent progenitors to a given lineage, followed by the maturation of the committed cells. To study the transcriptional events controlling these processes, we have investigated the role of C/EBP proteins in lineage choice of multipotent hematopoietic progenitors (MEPs) transformed by the E26 virus. We found that forced expression of either the alpha or beta isoforms of C/EBP in MEPs induced eosinophil differentiation and that in addition, C/EBPbeta could induce myeloid differentiation. Conversely, dominant-negative versions of C/EBPbeta inhibited myeloid differentiation. C/EBP-induced eosinophil differentiation could be separated into two distinct events, lineage commitment and maturation. Thus, eosinophils induced by transactivation-deficient C/EBPbeta alleles were found to be blocked in their maturation, whereas those expressing wild-type C/EBP proteins were not. Likewise, a 1-day activation of a conditional C/EBPbeta allele in multipotent progenitors led to the formation of immature eosinophils, whereas sustained activation produced mature eosinophils. These results show that C/EBP can induce both myeloid and eosinophil lineage commitment and that transactivation independent and dependent C/EBP functions are required during eosinophil lineage commitment and maturation, respectively.

Figure 1

Expression of C/EBPα and C/EBPβ isoforms in chicken hematopoietic cell types. Twenty micrograms of total RNA extracted from HD3, HD37 (erythroblast), HD50 (MEP), 1A1 (eosinophil), HD13 (promyelocyte), Bm2c2 (monoblast), HD50M (myeloblast), HD11 (macrophage), RPL12 (B-cell), and MSB-1 (T-cell) were probed with chicken C/EBPα, C/EBPβ, and GAPDH cDNAs. Because of comigration with 18S RNA, the C/EBPβ signal appears as a doublet.

Figure 2

Effect in MEP cells of chicken C/EBP isoforms on myeloid and eosinophil differentiation. (A) Structure of recombinant E26-C/EBP proviruses. (LTR) Long terminal repeat; (Θ) packaging signal; (IRES) internal ribosomal entry site. (B) E26-WT, E26-C/EBPα, E26-C/EBPβ, and E26-βD63N viruses were generated and used to infect blastoderm cells by cocultivation with the Q2bn packaging cells. Transformed colonies were isolated 2 weeks later, expanded in liquid culture for ∼2 weeks more and phenotyped by FACS analysis. Histograms show the proportion of cells positive for MEP21 (MEP cells), EOS47 (eosinophils), and MYL51/2 (myeloblasts) in each single colony-derived population. For each virus, clones are ordered according to decreasing MEP21 expression (left to right).

Figure 3

Cooperative activation of the EOS47 promoter by Ets-1 and chicken C/EBPs. One microgram of reporter plasmid (EOS47/−152-LUC) and 250 ng internal control plasmid (pRSV–βGal) were cotransfected into Q2bn fibroblasts along with 100 ng expression vectors for c-Ets-1 (pCRNCM-c-Ets-1) and 10 ng and 30 ng CMV expression vectors for C/EBPα, C/EBPβ, and C/EBPβD63N (increasing left to right). Empty expression vector was added to achieve equivalent DNA concentrations in all transfections. Luciferase activity was determined 48 hr later and normalized to the β-galactosidase activity. The basal level obtained with pEOS/−152-LUC alone was arbitrarily assigned a value of 1. The values shown represent the average of three or more experiments with similar results.

Figure 4

Characterization of eosinophils transformed by E26 viruses coexpressing various C/EBP forms. (A) FACScan profiles of EOS47 expression (left) and peroxidase staining (right) of representative eosinophil-rich clones transformed by E26–C/EBPα, E26–C/EBPβ, and E26–βD63N viruses. Peroxidase positive cells contain dark granules. (B) Immunofluorescence (left) and combined peroxidase and DAPI staining (right) of the same field of EOS47 FACS-sorted normal chick bone marrow cells. Peroxidase-positive cells are indicated by arrows.

Figure 5

Characterization of dominant negative mutants of C/EBPβ. (A) Structure of dominant-negative rat C/EBPβ mutants. (*) Position of the RK mutation. The sequences show a comparison between the basic regions (underlined) of rat C/EBPα and C/EBPβ, and the position of the substitutions introduced in the C/EBPβ/RK mutant (arginine-238 and lysine-239 into alanines). The numbers indicate the positions of the first and last amino acid shown for each protein. (B) Capacity of mutant C/EBPs to block the transactivation of the mim-1 promoter by C/EBPα. The ability of the βΔTA1, βΔTA2, and βΔTA2/RK alleles to inhibit C/EBP dependent transactivation was tested on the mim-1 promoter. One microgram of pMim-1ΔC–LUC was cotransfected with 0.25 μg of pRSVβGal, 0.5 μg pCMV–C/EBPα (chicken C/EBPα expressed from the CMV promoter). Two and five micrograms of CMV expression vector for the dominant-negative alleles were cotransfected as indicated. As a specificity control, 5 μg of a CMV expression vector for the I(16) protein (a dominant-negative inhibitor of the AML complex) was used. Relative transactivation values were calculated as in Fig. 3; the activity of pMim-1ΔC–LUC in the presence of C/EBPα was arbitrarily assigned a value of 1.

Figure 5

Characterization of dominant negative mutants of C/EBPβ. (A) Structure of dominant-negative rat C/EBPβ mutants. (*) Position of the RK mutation. The sequences show a comparison between the basic regions (underlined) of rat C/EBPα and C/EBPβ, and the position of the substitutions introduced in the C/EBPβ/RK mutant (arginine-238 and lysine-239 into alanines). The numbers indicate the positions of the first and last amino acid shown for each protein. (B) Capacity of mutant C/EBPs to block the transactivation of the mim-1 promoter by C/EBPα. The ability of the βΔTA1, βΔTA2, and βΔTA2/RK alleles to inhibit C/EBP dependent transactivation was tested on the mim-1 promoter. One microgram of pMim-1ΔC–LUC was cotransfected with 0.25 μg of pRSVβGal, 0.5 μg pCMV–C/EBPα (chicken C/EBPα expressed from the CMV promoter). Two and five micrograms of CMV expression vector for the dominant-negative alleles were cotransfected as indicated. As a specificity control, 5 μg of a CMV expression vector for the I(16) protein (a dominant-negative inhibitor of the AML complex) was used. Relative transactivation values were calculated as in Fig. 3; the activity of pMim-1ΔC–LUC in the presence of C/EBPα was arbitrarily assigned a value of 1.

Figure 6

Effects of dominant-negative rat C/EBPβ mutants on MEP differentiation. Two-day blastoderm cells were infected by cocultivation with Q2bn cells transfected with E26–WT, E26–βΔTA1, E26–βΔTA2, and E26–βΔTA2/RK DNAs. Transformed colonies were expanded and phenotyped by FACS analysis. The percentages of MEP21-, EOS47-, and MYL51/2-positive cells for each clone analyzed is shown (the clone order is identical in each of the three columns).

Figure 7

Effects of activating chicken C/EBPβ in MEPs transformed by E26–βER virus. (A) Structure of the E26–βER virus. (B) Western blot analysis of C/EBPβ–ER protein in E26–βER MEPs (βER cl.1 and cl.2; lanes 4,5) and of endogenous C/EBPβ in E26-WT-transformed myeloblasts (lanes 2,3). An E26–WT control MEP clone is shown for comparison (WT cl.1; lane 1). Western blotting was done with anti-chicken C/EBPβ antiserum on equal protein amounts from each clone. Bands corresponding to the endogenous C/EBPβ and exogenous C/EBPβ–ER are indicated. (C) Down-regulation of MEP21 expression by C/EBPβ–ER activation. E26–WT and E26–βER-transformed MEPs were treated with β-estradiol or mock treated, and their expression of MEP21 antigen, (D) EOS47 antigen, and (E) MYL51/2 antigen measured by FACS analysis after 2, 5, 8, and 12 days. Data from two representative clones of each type are shown. (▪) E26–WTcl.1; (♦) E26–WTcl.2; () E26–βERcl.1; (⋄) E26–βERcl.2.

Figure 8

Effect of transient β-estradiol treatment on E26–βER-transformed MEP clones. (A) Experimental protocol. Cells were treated for 1–4 days with β-estradiol, thoroughly washed, and incubated in the absence of the hormone for a total of 8 days, at which time they were subjected to FACS analysis. Changes in MEP21 (B), EOS47 (C), and MYL51/2 (D) expression in MEP clones transformed by E26–WT and E26–βER after transient exposure to β-estradiol for 1–4 days (as indicated) were determined by antibody staining and FACS analysis at day 8 after initiation of the experiment. (Black line) E26–WTcl.1; (blue line) E26–βERcl.1; (red line) E26–βERcl.2. (E) May–Gruenwald–Giemsa (left) and peroxidase staining (right) of untreated E26–βER cl.1 cells (a,b), or cells treated for 1 day (c,d), or 2 days (e,f) with β-estradiol. Note that 30% of the cells in c and d are EOS47 positive.

Figure 9

Model for the role of PU.1 and C/EBP in lineage commitment (A) and in eosinophil maturation (B). As explained in the text, C/EBPβ, although predominantly inducing eosinophil differentiation, may, in addition, be partially redundant with PU.1 with respect to induction of myeloid lineage commitment.