Vertebrate limb development (L15) Flashcards
Where do limb buds form?
Limb buds appear first as protrusions from the flank at precise positions along the AP axis of the embryo (limb fields). The determination of where these are is an intrinsic property of the lateral mesoderm. We know this because you can graft the area of the prospective limb into another area where there shouldn’t be one and it will start to form one (doesn’t respond to surrounding signals) Also, if you graft a non-limb area of the flank into a limb area it will not start to form a limb.
What specifies limb identity?
Two T-box transcription factors specify limb identity. You can use in situ hybridisation to visualise the restricted expression pattern of the genes that control fore and hind limb development. Tbx5 is expressed in forelimb bud and Tbx4 is expressed in the hind limb bud. We know Tbx5 can block expression of Tbx4 because if you introduce a construct expressing Tbx5 in the hind limb bud, the expression of Tbx4 is turned off. It was thought that these genes were the only ones that would give the identity of limbs due to fain of function studies, where you switch the expression of the T-box genes, leading to a reversal of limb identity. However, this is only true in the chick - In a mouse, if you do a loss of function of Tbx4, you still get a hind limb. Therefore, it appears that another gene (Pitx1) in the hind limb bud is the one actually driving limb identity - If you KO this you get hindlimb identity and Tbx4 expression. Also, if you force its expression in the forelimb, you get the expression of Tbx4 and hind limb identity.
How do Hox genes contribute to limb formation?
It is thought that limb fields form specific areas because of Hox proteins which allows for retinoic acid to be produced, causing Tbx TFs. this may also just be Hox genes acting directly on Tbx - its unclear. The boundary of Ho6 marks where the forelimb is going to form. You can also replace where the limb is going to form using ectopic expression of an FGF soaked bead - this leads to an ectopic limb.
What is the current model for the initiation of limb bud formation?
Hox proteins -> RA -> Tbx TFs -> Fgf10 (intermiediate mesoderm) -> Fgf8 (ectoderm (lateral plate))
Wnt signalling restricts FGF10 to just the limb bud. Wnt also causes expression of FGF8 in the apical ectodermal ridge in the limb bud. The Tbx TF will chnage whether the limb bud is front or back. FGF8 is generated in the intermediate mesoderm, this drives expression of FGF10 in the lateral plate. Wnt signalling restricts the expression domain of fgf10 in the lateral plate. Wnt signalling restricts the expression domain of Fgf10, so it is concentrated just in the limb bud. Wnt also causes expression of FGF8 in the apical ectoderm ridge in the limb bud (so the fgf8 from the intermediate initiates fgf10 in the lateral plate mesoderm - the wnt2b localised it to one region, then the fgf10 initiates wnt3a in the ectoderm next to the region - this is where the limb bud forms
What are the 3 main regions of the limb bud?
The apical ectodermal region, the progress zone and the zone of polarising activity. The PZ is made of mesenchymal cells
What are the actions of the AER?
The AER is necessary for the PZ (limb growth). More distal skeletal elements are • specified as the limb growth.
If you remove the AER, the earlier you do so, the less distal structures you develop. This lead to the suggestion of the 1st model for proximo-distal patterning:
That the AER is acting on the PZ to drive outgrowth by promoting proliferation of the cells. Therefore, the less time spent in the progress zone, the more proximal the structure – however this model isn’t correct, because it was demonstrated using fate maps that there are already progenitors that can contribute to proximal and distal structures.
Explain the 2 signal model of proximo-distal patterning
Where it is suggested that there are 2 signals involved
You initially have progenitors that can contribute to all structures in the limb, but the fate they adopt is dependent on how much of each of the antagonistic signal they receive. Obviously, the distal signal is the FGF coming from the AER. Then the proximal signal is RA
This interpretation of positional information is conserved in flies and vertebrates. – A similar antagonistic system is seen in the leg disc of a fly (Distaless is homotogue of Hox and Homothorax is the homolog of Meis)
We know that at different places along the gradients we get expression of different genes. For example , where retinoic acid levels are high, we get expression of Meis. We know Meis is controlled by RA, because if you KO the gene in a mouse that makes the RA, you lose Meis expression and proximal structures. This is the same with Hox 11, Hox 13 and FGF
How does the ZPA contribute to limb patterning?
The ZPA controls AP patterning
This is seen when an anterior graft of the ZPA creates a mirror image duplication of autopods (fingers) (if done early enough, there is also duplication of the ulna) – the same thing also occurs if you use Shh soaked beads
Wolperts French flag model also applies to the ZPA
Therefore, AP patterning is controlled by a morphogen gradient.
The phenotype of Shh mutant limb is consistent with the morphogen gradient hypothesis.
I.e. in the absence of Shh, there is a complete loss of the distal most skeletal elements. Also, there is loss of the zeugopods. You still get the 1st digit because it doesn’t require Hh signalling. (It is a the lowest conc bit of the gradient). However, you get a defect in outgrowth as well due to the coordination between the different axes.
There is coordination between proximo-distal and AP patterning
ZPA helps maintain FGF expression (so without this you get problems with outgrowth)
How is the limb bud patterned in the dorsoventral axis?
• There are 3 genes with essential roles in DV patterning.
They are Wnt7a (on the dorsal side), En1 (on the ventral side), and Lmx1b (divides the limb along the DV axis)
Like the AP patterning, this is also conserved in the drosophila wing disc. Apterous is a homolog of Lmx1b and controls dorsal fate.
Conditional KO of Lmx1b causes a ventrilisation of the limb. - I.e. instead of forming fur, you get signs of a paw forming on both sides.
A proposed model for this control is as follows.
So, you have BMP signalling take place in the ventral part of the limb, which drive EN1 expression (again, only ventrally). En1 inhibits Wnt7a, which activates Lmx1b. So, where En1 is present, you get no dorsal fates, because Wnt is blocked. This limits the Lmx1b expression to the dorsal limb bud.
If you lose Lmx1b, you get all ventral fates, suggesting this is the default for the cells.
