Relationship of the Dorsoventral Pigment Boundary to Lineage Compartments and the Lateral Somitic Frontier
The ectodermal notch that we used to mark the boundary between embryonic dorsum and embryonic flank is a characteristic feature in vertebrate embryos. In cell lineage studies carried out in the chick system, the notch serves as a landmark for the boundary between dermis derived from somitic mesoderm and dermis derived from lateral plate mesoderm and has been termed the “lateral somitic frontier” (Olivera-Martinez et al. 2000; Sudo et al. 2001; Burke and Nowicki 2003; Nowicki et al. 2003). Although fate-mapping studies have not been carried out in mammalian embryos, somite- and lateral plate-derived mesoderm could give rise to precursors for dermis dorsal and ventral to the limb–body wall junction, respectively. However, this notion conflicts with our observation that the future pigmentation boundary lies ventral to the ectodermal notch (see Figure 7).
To examine directly the relationship between the pigmentation boundary and dermis derived from lateral plate mesoderm, we made use of a Cre transgene driven by the Hoxb6 promoter that was developed by Kuehn and colleagues (Lowe et al. 2000). As described by Lowe et al. (2000), midgestation embryos carrying both the Hoxb6-Cre transgene and the R26R lacZ reporter gene (Soriano 1999) exhibit X-Gal staining in lateral plate mesoderm but not somite-derived mesoderm of the trunk. In whole-mount skin preparations from P1.5 or P4.5 neonatal animals, we observed a ventral band of dark X-Gal staining corresponding to lateral plate-derived dermis, which represents 63% of the total circumference (Figure 8A). However, in parallel preparations from at/at mice, the ventral pheomelanin domain represents 47% of the total skin circumference; therefore, the proportions of total skin circumference occupied by dorsal eumelanin and somite-derived dermis are 53% and 37%, respectively (Figure 8B). These results indicate that the pigmentation boundary is clearly distinct from, and more ventral to, the boundary between lateral plate- and somite-derived dermis.
Figure 8  Comparison of the Dorsoventral at/at Pigmentation Boundary to the Lateral Somitic Frontier
(A) Dorsoventral slices of skin from at the midtrunk region prepared such that the dorsal midline lies in the center of the slice. Sections were taken at P1.5 (a) or P4.5 (b–e) from at/at or R26R/+; Tg.Hoxb6-Cre/+ mice (the latter were stained with X-Gal), as described in Materials and Methods. For purposes of comparison, images were proportionally scaled. The boundary of X-Gal staining marks dermis derived from lateral plate versus dermis derived from mesoderm (the lateral somitic frontier) and lies more dorsal than the at/at pigmentation boundary.
(B) Quantitation of mean (± SEM) dorsal pigmentation area (n = 5) and somite-derived dermis area (n = 3) shows a significant difference (p < 0.005, t-test).
(C) RNA in situ hybridization showing that Tbx15 expression at E11.5 is complementary to En1 expression on the flank (scale bars = 200 μm). The arrow indicates the boundary between the expression domains of the two genes. Because the pigmentation boundary lies in register with the limb–body wall junction (see Figure 2), we wondered whether mechanisms used for dorsoventral limb patterning might be related to those used to establish the pigmentation boundary. In the developing limb, Engrailed1 (En1), Wnt7a, and Lmx1b are part of a network whose restricted domains of expression help to establish dorsoventral identity (reviewed in Niswander 2003). En1 is transiently expressed in the developing flank; at E11.5, transverse abdominal sections reveal domains in the neural tube, somite-derived mesenchyme, and the ventral body wall (Figure 8C). An adjacent section hybridized with Tbx15 reveals a complementary pattern in the flank, which provides additional evidence for developmental mechanisms that establish a pigmentation boundary entirely within lateral plate mesoderm and independent of lineage restrictions imposed by the lateral somitic frontier.