GATA1 Changes DNA-Binding Fashion in a Binding-Site-Specific Manner and Alters Transcriptional Activity during Erythropoiesis

GATA1 is a transcription factor that coordinately regulates multiple target genes during the development and differentiation of erythroid and megakaryocytic lineages through binding to GATA motif (A/T)GATA(A/G). GATA1 has four functional domains, i.e., two transactivation domains reside in amino- and carboxyl- terminus, which transactivate GATA1 target genes redundantly and/or cooperatively, and two zinc-finger domains in the middle of the protein. The two zinc finger domains of GATA1 have been characterized extensively and their links to human diseases have also been identified. Carboxyl-terminal side zinc (C)-finger is essential for the DNA binding of GATA1, whereas amino-terminal side zinc (N)-finger retains insufficient binding activity to the GATA motifs by itself, but contributes to stabilize the binding of C-finger to a double GATA site arranged in a palindromic manner. Of note, while this two-finger structure is conserved in six distinct vertebrate GATA factors, there exist GATA factors with single zinc finger in non-vertebrates, indicating that only the C-finger and following basic tail region are evolutionary conserved in both vertebrate and non-vertebrate GATA factors. In our transgenic rescue analyses, GATA1 lacking the N-finger (ΔNF-GATA1) supports, if not completely, the erythropoiesis in mice, but mice without C-finger (ΔCF-GATA1) die in utero showing similar phenotype to the mice with complete loss-of-GATA1-function. Therefore, roles that the N-finger plays have been assumed to be evolutionally acquired features during molecular evolution.

In this study, we have examined GATA-motif configuration-specific modulation of GATA1 function by using composite GATA elements in which two GATA motifs aligned side-by-side, either tandem or palindromic. We have defined changes in the GATA1 binding and transactivation activity in accordance with the arrangement of cis-acting GATA motifs. While GATA1 binds to Single-GATA in a monovalent way via C-finger without the influence of N-finger, the N-finger appears to contribute to specific bivalent binding of GATA1 to Pal-GATA, i.e., the N- and C-fingers in a single GATA1 molecule individually bind to two GATA motifs aligned in a palindromic orientation. Showing very good agreement with the human case analyses, the transgenic expression of G1^R216Q that lacks N-finger-DNA interaction potential hardly rescues the GATA1-deficient mice due to defects in definitive erythropoiesis, indicating that roles owed by R216 residue are vital for the GATA1 activity in vivo. The N-finger also contributes to GATA1 homodimer formation, which is a prerequisite for two GATA1 binding to two GATA motifs aligned in a tandem orientation. Each GATA1 C-finger in the dimeric GATA1 protein binds to each GATA motif in Tandem-GATA. In this regard, we previously found in a transgenic complementation rescue assay that mutant GATA1 molecule G1^3KA, which lacks the dimerization potential but possesses most of the other N- and C-finger functions, hardly rescues the GATA1-deficient mice from embryonic lethality, indicating that the GATA1 dimerization is important to attain full GATA1 activity. We surmise based on these observations that the configuration of cis-acting GATA motifs located in the regulatory regions of the GATA1 target genes critically influences the DNA-binding of GATA1 and controls transcription of the genes.