Fig. 4. The midline phenotype in the Shh mutant differs from that in the Bmpr1a;Bmpr1b mutant mouse. TUNEL analysis (A,B) and RNA in situ hybridization (C-H) on E10.5 coronal sections. (A,B) TUNEL-positive cells are detected at the most dorsomedial neuroepithelium in the Shh mutant (B, arrowhead), where apoptosis occurs in control embryos (A, arrowhead). (C-H) The dorsal midline markers Msx1 (C,D) and Bmp4 (E,F) are expressed in the Shh mutant (arrowheads) and Foxg1 remains excluded from that area (H, arrowheads). Scale bar: 200 µm. (I) Model illustrating distinct mechanisms of HPE. In the wild type (left), BMPs acting downstream of GLI3 (Grove et al., 1998; Kuschel et al., 2003; Theil et al., 1999) and ZIC2 are both required for the formation of the dorsal midline (ZIC2 is also required ventrally). Note that Fgf8 may also regulate dorsal midline development (Crossley et al., 2001), although only reduction, not loss, of Fgf8 expression leads to dorsal HPE (Storm et al., 2006). SHH acts indirectly to generate ventral cells by antagonizing GLI3, which in turn relieves the repression of Fgf expression (Aoto et al., 2002; Gutin et al., 2006; Rallu et al., 2002). In the Bmpr double mutant (middle), the dorsal midline fails to form despite the maintenance of Zic2 expression. However, ventromedial cell fates are not affected, as is also the case in the MIH subclass of HPE. In the Shh mutant and other mutants in which SHH signaling is disrupted (right), dorsal midline features are initiated but ventromedial cells are lost.