CCAAT/Enhancer-Binding Protein ε27 Antagonism of GATA-1 Transcriptional Activity Is Mediated by a Unique N-Terminal Repression Domain, Is Independent of Sumoylation, and Does Not Require but Is Enhanced by DNA-Binding.

Abstract

The CCAAT/enhancer binding protein epsilon (C/EBPε) is critical for the terminal differentiation and lineage-specific gene expression of granulocytes, and expression of C/EBPε32 and its shorter 27 kD and 14 kD isoforms is developmentally regulated during neutrophil granulocyte differentiation. We have defined a novel role for the unique 27 kD isoform (C/EBPε27) as a potent antagonist of GATA-1-mediated transactivation of the promoter of the gene encoding the eosinophil secondary granule protein, major basic protein (MBP) (Du et al, J. Biol. Chem. 2002; 277:43481–43394). We also showed that these two transcription factors physically interact in eosinophil cell lines in vivo. In the present studies, we performed the first structure-function analyses of the C/EBPε27 isoform to map its potent repressor domains, with comparisons to the C/EBPε32 and C/EBPε14 isoforms, using transactivation assays of the MBP P2 promoter in the presence of GATA-1. Our results show that the repression of GATA-1 is mediated in part by the unique N-terminus of C/EBPε27 (not shared with other C/EBPε isoforms) in combination with part of a previously identified RDI domain (shared with full length C/EBPε32). We show further that this repressor activity is independent of DNA binding (via deletion of the basic region of C/EBPε27) as well as of sumoylation of the RDI “VKEEP” sumoylation consensus site present in both the C/EBPε32 and C/EBPε27 isoforms, and conserved in the C/EBPε proteins of many other species. Thus, our findings identify the unique N-terminus of the C/EBPε27 isoform, a distinct 68 amino acid sequence not shared with any other C/EBPε isoforms or other C/EBP family members, as the minimum repressor domain required for potent antagonism of GATA-1 activity. Of interest, fusion of this novel 68 amino acid sequence to the N-terminus of full length C/EBPε32 converted it into a partial repressor of GATA-1, but did not alter the transactivation potential of the C/EBPε32 isoform itself. The mechanism for maximal C/EBPε27 attenuation of GATA-1 activity requires a combination of both GATA-1-C/EBPε27 protein-protein interaction and C/EBPε27 binding to the proximal C/EBP consensus site immediately upstream in the target promoter. Neither C/EBPε32 nor C/EBPε14 inhibited C/EBPε27 antagonism of GATA-1, supporting a protein-protein interaction mechanism for its repressor activity that is enhanced by, but does not require, DNA binding to a proximal C/EBP site. Expression of the C/EBPε27 isoform likely serves to titrate and/or turn off expression of secondary granule protein genes such as MBP during eosinophil terminal differentiation, when these genes are ultimately silenced in the mature cell. These studies illustrate the unique regulatory (activating versus repressor) activities for the various C/EBPε isoforms, activities consistent with their developmentally regulated expression and lineage-specific activities during granulocyte (both neutrophil and eosinophil) differentiation.