A cascade of transcriptional regulators may control dorsoventral development of the Drosophila embryo. Genetic analysis showed that two zygotic loci, twist and snail, are required for the formation of the mesoderm. In mutant embryos of twist and snail, the cells on the ventral side do not invaginate and no ventral furrow is formed resulting in embryos lacking all derivatives of the mesoderm. Previously, a novel homeobox gene cluster at the 93E1-5 region of the third chromosome was discovered. The expression pattern of NK-4 and NK-3 homeobox genes, together with that of the twist gene during embryogenesis suggested that these genes may regulate each other and, thereby, may control precise mesodermal cell differentiation. We demonstrated that the NK-4 protein has DNA binding activity to a known binding site for Ubx protein using a gel-shift assay with in vitro translated NK-4 protein. Domain analysis of the NK-4 protein, which was revealed by transient transfection with various mutant constructs containing different parts of NK-4 protein fused to the GAL-4 DNA binding domain, showed that the amino-terminal part of NK-4 protein is a transcriptional activator domain. Extracts from myoblast cells transfected with NK-4 expression vector and reporter plasmids containing the CAT gene driven by the NK-3 promoter and upstream region (-1594 to +25) showed a 5-fold increase in CAT activity compared to that from cells transfected with control plasmids alone, suggesting that NK-4 is a positive activator of the NK-3 homeobox gene. This result was confirmed by an in vitro transcription assay using nuclear extracts from the Drosophila embryo. Interestingly, it appeared that NK-3, itself, showed negative repressor activity of its own promoter. We have cloned the upstream region of NK-4 and sequenced 3.5 kb of DNA. We found two clusters of E-boxes (CANNTG) which are known to be binding sites for HLH proteins such as MyoD and twist. Each cluster has three E-boxes within 35 nucleotides. We demonstrated that embryonic nuclear extracts showed binding activity to these E-boxes by a gel-shift assay. We have also cloned the twist gene by PCR and expressed the twist protein in E. coli. Studies of the regulation of NK-4 by twist are underway.
