4 RA Signaling in Limb Development and Regeneration …
109
RA is generated in the somites showing striking uniformity, and there are some
downstream genes which emerge in common such as the meis genes, fgfs, fibin and
of course the hox genes. Clearly there is much to be done to look for even more
commonality, particularly in the case of the amphibian limb bud. Almost nothing is
known about the latter because the amphibian limb bud is a grossly under-studied
model.
Is RA a Morphogen?
The early studies suggested that RA would be the morphogen of the ZPA that organizes the AP axis. However, this concept has been abandoned subsequent to several
observations:
• RA addition induces RARβ2 and hoxb-8 expression, while a ZPA graft does not
(Noji et al. 1991)
• shh, which is released from the ZPA as an AP morphogen, is induced by RA (Noji
et al. 1991; Wanek et al. 1991; Tamura et al. 1993; Riddle et al. 1993)
• RA acts through the AER (via fgf-4) to generate ZPA-like activity, whereas the
ZPA itself does not (Niswander et al. 1994; Helms et al. 1996)
• The organization of the PD axis and the proliferation and distalisation of limb cells
requires inhibition of RA signaling via FGFs (Rosello-Diez et al. 2011; Cooper
et al. 2011; Mercader et al. 2000).
Emerging Uniformity
What emerges from the data described above is a striking uniformity across development and regeneration and across vertebrate limbs and fins (summarized in Figs. 4.4
and 4.5). RA is required for limb development and regeneration, and then, for outgrowth and specification of all three axes (AP, PD and VD) as it posteriorizes, proximalizes and ventralizes limbs. In the absence of RA, forelimbs are not formed,
whereas hindlimb formation does not seem to require RA (although this has not
been investigated across all vertebrates). In contrast, excess RA duplicates limbs
when applied very early, but causes teratogenic effects when applied later.
There is no reason to expect a priori that the molecular mechanisms and gene
pathways that govern limb formation should be identical for development and regeneration across species, and indeed, they do not seem to be (Fig. 4.5). In amphibian
development, RA is present throughout the mesenchyme of the forelimb bud, but
in regeneration, it is present in the epithelial cap. The role of the meis genes in PD
specification looks clear in chick development and axolotl regeneration, but not so
in mouse development. The distribution of the metabolic and catabolic enzymes is
different between amniotes and anamniotes (Fig. 4.4). Some of the differences are
likely due to structural differences in the way the limbs develop (as for example, RA
109
RA is generated in the somites showing striking uniformity, and there are some
downstream genes which emerge in common such as the meis genes, fgfs, fibin and
of course the hox genes. Clearly there is much to be done to look for even more
commonality, particularly in the case of the amphibian limb bud. Almost nothing is
known about the latter because the amphibian limb bud is a grossly under-studied
model.
Is RA a Morphogen?
The early studies suggested that RA would be the morphogen of the ZPA that organizes the AP axis. However, this concept has been abandoned subsequent to several
observations:
• RA addition induces RARβ2 and hoxb-8 expression, while a ZPA graft does not
(Noji et al. 1991)
• shh, which is released from the ZPA as an AP morphogen, is induced by RA (Noji
et al. 1991; Wanek et al. 1991; Tamura et al. 1993; Riddle et al. 1993)
• RA acts through the AER (via fgf-4) to generate ZPA-like activity, whereas the
ZPA itself does not (Niswander et al. 1994; Helms et al. 1996)
• The organization of the PD axis and the proliferation and distalisation of limb cells
requires inhibition of RA signaling via FGFs (Rosello-Diez et al. 2011; Cooper
et al. 2011; Mercader et al. 2000).
Emerging Uniformity
What emerges from the data described above is a striking uniformity across development and regeneration and across vertebrate limbs and fins (summarized in Figs. 4.4
and 4.5). RA is required for limb development and regeneration, and then, for outgrowth and specification of all three axes (AP, PD and VD) as it posteriorizes, proximalizes and ventralizes limbs. In the absence of RA, forelimbs are not formed,
whereas hindlimb formation does not seem to require RA (although this has not
been investigated across all vertebrates). In contrast, excess RA duplicates limbs
when applied very early, but causes teratogenic effects when applied later.
There is no reason to expect a priori that the molecular mechanisms and gene
pathways that govern limb formation should be identical for development and regeneration across species, and indeed, they do not seem to be (Fig. 4.5). In amphibian
development, RA is present throughout the mesenchyme of the forelimb bud, but
in regeneration, it is present in the epithelial cap. The role of the meis genes in PD
specification looks clear in chick development and axolotl regeneration, but not so
in mouse development. The distribution of the metabolic and catabolic enzymes is
different between amniotes and anamniotes (Fig. 4.4). Some of the differences are
likely due to structural differences in the way the limbs develop (as for example, RA