Fig. 3. Drosophila Shn and its vertebrate homolog human Shn1 can mediate Bmp-responsive transcriptional activation through the BRE. (A) Structural organization of Shn, hShn1, hShn2 and hShn3. Shn contains a total of eight zinc fingers (vertical bars) with characteristic spacing. All zinc finger domains are of the C2H2 type, except for finger 3, which is a C2HC type. All three human proteins contain the first and second set of paired zinc fingers (1/2 and 4/5 in Shn), but lack the triplet set near the carboxy terminus (6/7/8). The C2HC finger is absent from hShn2. Shn and hShn1 share 73% identity in the 1/2, and 87% in the 4/5 paired finger domains, whereas finger 3 shows only 27% identity. The fly and human Shns have minimal identity in the remainder of the protein. By contrast, hShn1, hShn2 and hShn3 share 26-31% sequence identity outside of their finger regions. (B) Molecular phylogeny of Shn proteins from D. melanogaster, C. elegans, X. tropicalis, M. musculus and H. sapiens generated using CLUSTALW. Invertebrate Shn proteins most closely resemble vertebrate Shn1. (C) In Xenopus animal cap assays, hShn1 stimulates wild-type Xvent2-BRE reporter gene expression even in the absence of CABR (i.e. in the presence of endogenous Bmp signaling). However, co-expression of hShn1 together with CABR results in a stronger induction of the reporter. (D) Drosophila Shn only weakly induces the Xvent2-BRE reporter in Xenopus animal caps. However, in the presence of CABR, Shn shows a strong induction of the reporter. Results are presented as fold activation relative to basal activity of the wild-type reporter in the absence of Shn proteins or CABR.