Introduction Xenopus Clements et al., 2001; Clements and Woodland, 2003; Engleka et al., 2001; Hudson et al., 1997; Yasuo and Lemaire, 1999 Xanthos et al., 2001 Xnr4 Yasuo and Lemaire, 1999 Xsox17 Mix.1 Mixer Sox17 Xenopus laevis Xsox17α 1 α 2 β Xsox17 Hudson et al., 1997; Zorn et al., 1999 Xsox17 Xsox17 Clements and Woodland, 2000; Hudson et al., 1997 Clements et al., 2003 Sox17 Kanai-Azuma et al., 2002 Casanova Sox17 Aoki et al., 2002 Sox17 Casanova Xsox17 Xsox17 Xenopus Casanova Xenopus Kikuchi et al., 2001; Woodland and Clements, 2003 Xenopus Endodermin Hnf-1β Clements et al., 2003 Sinner et al., 2006 Xsox17 Xsox17α 1 Materials and methods Biological materials X. laevis Wilson et al., 1986 3 3 2 2 Transgenic methods Kroll and Amaya (1996) Transient transgenesis in embryos and oocytes Renilla Renilla Renilla Xsox17α 1 Xsox17α 1 X. laevis gilli Xsox17α 1 Transgenic constructs and mutagenesis TGATCA CTAGTT ACTAGT CCATGG CCATGG TGATCA Sac1 Kpn1 Electrophoretic band-shifts (EMSAs) 32 2 TCACACCA ACGCACAT TCACACCT Chromatin immunoprecipitation (ChIP) 2 2 3 C3B1 F 5′ GCCAATAGACACCTTTCTAG 3′ C3B1 R 5′ GAGAATGGGACTGTGTTAAC 3′ Xsox17α ORF F 5′ GGACGAGTGCCAGATGATG 3′ Xsox17α ORF R 5′ CTGGCAAGTACATGTGTCC 3′ Xom Promoter F 5′ TGTTGGCTGAGTAGGAATGAGAGG 3′ Xom Promoter R 5′ AGGCAGAGATCAGTACCACCT 3′ Messenger et al. (2005) Results Xsox17α 1 Xsox17α 1 X. laevis gilli Xsox17α 1 Xsox17β Supplementary Fig. 1 Xsox17β Xsox17 2 Xsox17α 1 Xsox17α 1 Fig. 1 Kroll and Amaya (1996) Figs. 1 Ahmed et al., 2004 Sasai et al., 1996 Fig. 1 Supplementary Tables 1 and 2 Supplementary Fig. 2 Clements et al., 2003 Hudson et al., 1997; Zorn and Mason, 2001 Supplementary Fig. 2 Zorn and Mason (2001) Xsox17α 1 Xsox17 Schlosser and Ahrens, 2004; Zygar et al., 1998 Xsox17 Xsox17β Supplementary Fig. 1 Xsox17α 1 Supplementary Fig. 2 Xsox17β Xsox17 2 Xsox17 Xenopus tropicalis X. tropicalis Xsox17α 1 Fig. 2 Fig. 2 Xenopus Ahmed et al., 2004; Latinkic et al., 2002 Dissection of the E-element Fig. 3 Fig. 3 Xsox17α 1 Figs. 3 Xsox17 Figs. 1 Xsox17 Clements et al., 1999; Yasuo and Lemaire, 1999 Ahmed et al., 2004 Transgenic analysis of the T-box and Sox sites in element B1 Fig. 3 Supplementary Fig. 3 Derrière White et al., 2002 Fig. 3 Fig. 3 Supplementary Table 2 Fig. 3 Transient transgenic analysis of B1 and C3 Xsox17 Wilson et al., 1986 Xsox17 Fig. 4 Fig. 4 White et al., 2002 Xsox17 Clements et al., 1999; Clements and Woodland, 2003; Yasuo and Lemaire, 1999 Fig. 4 White et al., 2002 White et al., 2002 Fig. 4 Xsox17 Clements et al., 1999; Clements and Woodland, 2003; Xanthos et al., 2001 Hemmati-Brivanlou and Melton, 1992 Fig. 4 Fig. 4 Xsox17 Sinner et al., 2004 Fig. 4 Sinner et al., 2004 Fig. 4 Hnf-1β Xsox17 Fig. 4 1 Fig. 4 1 Xsox17s Clements and Woodland, 2003 Mutational analysis of B1 shows that one Sox site is most important and it co-operates with the VegT site Fig. 5 Fig. 5 X. laevis Xsox17 Clements et al., 2003 Fig. 5 The B1 element is bound to Xsox17 in vivo Xsox17β Clements et al., 2003 Fig. 6 Xsox17α 1 Messenger et al., 2005 Xsox17β Xsox17β Discussion Xenopus Xsox17 Xsox17α 1 Xsox17 Introduction Xsox17α 1 Xsox17α 1 Activity of the C3 endodermal element Chen et al., 1997; Germain et al., 2000 Kofron et al. (2004) Xsox17 Xenopus Howell et al., 2002 Xsox17 C3 responds strongly to Activin, but with respect to vegetal expression and response to VegT, its behaviour is paradoxical. It is surprising that it does not respond to VegT in the embryo, both because C3 contains a consensus T-box core sequence and because VegT induces the expression of TGF-βs that, like Activin, act through Smads 2/3; indeed VegT depends on this signalling for its overall biological effect. The T-box site in C3 entirely overlaps a Fast1/Smad pair of sites, which may have a bearing on the fact that VegT actually inhibits C3 basal expression, just as its expression is inhibited in the vegetal pole compared to the animal. This is consistent with the observation that removal of the T-box site removes the inhibition (data not shown). C3 was identified by its ability to direct expression in the vegetal pole of transgenics, however here we simply scored expression in the vegetal pole itself, disregarding animal regions. Conversely, the luciferase measurements are of the ratio either of vegetal to animal expression or of VegT stimulated to control expression in the animal region. Therefore a high level of animal expression would mask vegetal activity of the promoter. However, it is important to note that both the VegT induction and vegetal expression are consistent. We believe that the response to TGF-βs is the important property of C3 and other problems are introduced by looking at small regulatory regions in isolation, where synergising and inhibitory effects are absent. Xsox17 Xsox17 The activity of the B1 endodermal element The activity of B1 is more straightforward. It is highly expressed in the presumptive endoderm in transgenics and it is much more highly expressed in the vegetal than the animal hemisphere in transient assays. It is also strongly stimulated by VegT. It contains a divergent T-box site which binds VegT in vitro and which mutation shows is partially responsible for the VegT stimulation in the embryo. Expression in oocytes shows that B1 responds directly to VegT through this site, although the stimulation is less than in embryos. This correlates with the fact that blocking TGF-β signalling in embryos with a truncated Activin receptor reduces VegT stimulation to about the oocyte level. Thus B1 responds directly to VegT and synergistically to other molecules that are downstream of the VegT-induced TGF-βs, which would principally be Nodal-related signals. These synergistic molecules are the Xsox17 proteins themselves in an autoregulatory loop. Sinner et al., 2004 Xsox17 Xsox17 Endodermin Ahmed et al., 2004 Davis et al., 1995 Endodermin Sinner et al., 2006 Xsox17 The kinetics of reporter expression suggest that the VegT and the T-box site are not simply needed at the very onset of endodermal gene expression in the initiation phase, but that this continues into gastrulation during the establishment phase. It is provocative that the Sox site of principal importance in B1 is very near to the VegT site, suggesting a direct interaction between VegT and Xsox17, although we have been unable to detect this by immunological co-precipitation. Xsox17 Xsox17 Xsox17 Clements et al., 1999; Clements and Woodland, 2003; Wardle and Smith, 2004; Wylie et al., 1987; Yasuo and Lemaire, 1999 Xsox17 HNF1-β Endodermin Gata5 Clements et al., 2003; Sinner et al., 2006 Xsox17 Clements et al., 2003 Xsox17 Afouda et al., 2005 Xsox17 Zhang et al., 2005 Xsox17 Sinner et al., 2004 in vivo Xsox17 Clements and Woodland, 2003 Xsox17 Xsox17 Gurdon et al., 1993a,b Fig. 6 Mix/Bix Xsox17 Xsox17 Zhang et al., 2003, 2004 Xsox17 Sinner et al., 2006 Mangan and Alon, 2003 Mochizuki (2005) Appendix A Supplementary data Supplementary Fig. 1 Xsox17α 1 β  Supplementary Fig. 2 Xsox17α 1  Supplementary Fig. 3 Sequences of endodermal sub-elements B1 and C3. Possible Smad sites (italics), FoxH1 (synonym Fast1) sites (blue), Sox binding sites (green) and T-box half site (bold underline).  Supplementary Table 1 Typical transgenic experiment analyzing deletion mutants.  Supplementary Table 2 Analysis of normal transgenic mid-gastrula (stage 10.5) embryos and transgene expression in experiments analyzing the effects of endodermal element mutations.  Appendix A Supplementary data doi:10.1016/j.ydbio.2007.07.028