Student Reader

Limb development

What are the signaling centers controlling limb bud formation?

Apical Ectodermal Ridge
The Apical Ectodermal Ridge (AER) controls identity along the proximal/distal axis. The AER is a ridge of thickened ectoderm that forms along the entire edge of the limb bud at the dorsal/ventral boundary. This thickened ridge is analogous to ectodermal placodes in that each are areas of thickened ectoderm created by cell shape changes and they each signal to the underlying mesoderm. The mesoderm underlying the AER is called the Progress Zone.

RemovalThe AER was removed from the limb at various times. This resulted in limb truncations. Thus, the AER signals the underlying mesenchyme to allow proximal/distal outgrowth. Late removes results in only the most distal elements (ie, digits) missing; early removal results in more proximal elements also missing. Thus, proximal-distal patterning occurs progressively over time with proximal elements forming first.
HybridizationIn situ hybridization was used to determine what factors are expressed by the AER. Among those factors were FGF-4 and -8. To test if these are the factors responsible for AER's proximal-distal patterning activity, implantation of beads was performed in embryos stripped of their AER.
ImplantationThe AER was removed and beads soaked in FGF-8 were implanted. The implanted beads were nearly as effective as the AER itself. Thus, the AER is clearly a signaling center that secretes FGFs to induce underlying mesenchyme.

Zone of Polarizing Activity
The Zone of Polarizing Activity (aka ZPA or Polarizing Region) controls identity along the anterior/posterior axis (and thus digit identity). It is a mesenchyme located at the posterior end of the limb bud. Digit 1 is most anterior numeration proceeds posteriorly.

TransplantationA ZPA was grafted to an ectopic anterior position. This caused formation of a new anterior/posterior axis. This is similar to the grafting of the frog embryo DBL to an ectopic position to induce a new dorsal/ventral axis with a new neural tube and complete set of structures. The interpretation for the ZPA results is the same as in the DBL results: the grafted tissue must be a source of a secreted signal that acts upon neighboring cell.
HybridizationIn situ hybridization identified Shh as the potential signal secreted by the ZPA.
ImplantationImplanting a bead soaked with Shh at the anterior end of the limb will cause the limb to form with two anterio-posterior axes as though it were a reflection of itself.

Formation of structures along the anterio-posterior axis is dependent on the concentration of Shh. Naturally, the [Shh] is highest posteriorly where the ZPA is located. Adding a source of Shh at the anterior end will form a concentration gradient that is high anteriorly and posteriorly and intermediate in the middle. Mutations that lead to ectopic Shh are the most frequent cause of limb/digit duplication.

Progress Zone, Part I
The Progress Zone controls identity along the proximo-distal axis. The Progress Zone is the limb mesenchyme directly underneath the AER. The proximal/distal fate of a cell is determined by the amount of time it spends in the Progress Zone. Outgrowth of the limb is proximal→distal, meaning proximal structures develop first and distal structures last. In the early limb bud most mesenchymal cells are in the Progress Zone. However, as cells divide and the limb grows, more and more cells are found outside of the Progress Zone. They only begin to differentiate once outside the Progress Zone. This is similar to the nervous system, where progenitor cells only differentiate after leaving the ventricular proliferative zone.

LabelingDifferent cells are labeled with different markers. This reveals that as development proceeds, cells in the Progress Zone proliferate. As they proliferate, some cells are pushed out of the Progress Zone. Cells pushed out first become proximal structures. As limb development proceeds, cells i the Progress Zone continue to proliferate and additional cells are pushed out.

Progress Zone, Part II
Since the amount of time spent in the Progress Zone fates a cell along the proximal/distal axis, the labeling experiment explains why removing the AER at an early stage leads to greater limb truncations.

TransplantationAn early Progress Zone was transplanted onto a late limb bud. Many cells had already left the host's late Progress Zone. Grafting a new Progress Zone onto the tip of the limb caused a new population of cells to proliferate and leave the Progress Zone as the host cells had. The result is duplication of structures.
TransplantationAn early limb bud Progress Zone was removed and replaced with a late limb bud Progress Zone. Only the most proximal structures had time to leave the host Progress Zone. The donor Progress Zone was only left with the most distally determined cells. This resulted in a loss of intermediate limb elements.

The Apical Ectodermal Ridge (AER) produces FGF-4 and -8. The FGFs act on the Progress Zone and maintain the Polarizing Region (ZPA).

The Polarizing Region (ZPA) produces Shh, and this Shh gradient establishes the anterio-posterior axis. Furthermore, Shh acts on the Progress Zone and AER. Thus, Shh is required for proximal→distal outgrowth.

The Progress zone maintains the Apical Ectodermal Ridge (AER). The Progress Zone produces FGF10.

We have seen that the ZPA produces Shh and controls A/P identity.
The PZ and AER control P/D outgrowth.
The PZ produces FGF10 and the AER produces FGF8.

The new concept I want to introduce is that these signaling centers interact to maintain each other. For example, in addition to controlling A/P identity, Shh also acts on the PZ to to maintain its survival.
Similarly, the FGF10 produced by the PZ maintains the AER.
Finally, the FGF8 produced by the AER acts on both the PZ and the ZPA to maintain these structures.

1. Removal of AER stops formation of distal structures

2. Removal of AER causes loss of ZPA activity

3. Replacement of PZ mesoderm with flank
mesoderm causes degeneration of AER

4. Removal of ZPA causes loss of formation of distal structures (i.e. PZ activity and AER activity)

There are some experiments that tell us that the PZ, AER, and ZPA interact to maintain each other.

For example, we have seen that removal of AER blocks formation of distal structures. This suggests it is required to maintain the PZ.
Next, removal of the Aer also causes a loss of ZPA activity. Shh expression is lost. This shows the AER is required to maintain the ZPA.

The role of the PZ on maintaining the AER can be seen if we replace the PZ mesoderm with other flank mesoderm. This other mesoderm is unable to maintain the AER.

Finally, removal of the ZPA causes a loss in the formation of distal structures.

To review, as I mentioned earlier, the limb bud consists of mesenchymal cells arising from the lateral plate mesoderm, and an epithelial covering coming from the ectoderm. It is at this early stage of development, before any cell differentiation takes place, that different parts of the limb bud begin to produce signals that control development along the P/D and A/P axes. The basic point here is that these signaling sends transmit information to cells in the limb bud at an early stage of development, and it isn’t until much later that cells start to differentiate into different skeletal elements at specific locations, depending on the signals they received earlier.

How is limb bud development initiated?
  • Limb buds arrive at precise locations along the anterio-posterior axis via Hox genes. Hox gene expression patterns determine specific locations where FGF10 becomes expressed in the lateral plate mesoderm.
  • FGF10 is a signal that induces formation of the Apical Ectodermal Ridge. As soon as the AER forms, it immediately begins its function of controlling limb bud formation and distal outgrowth.
  • Later, Shh expression is induced by FGFs produced in the AER. Evidence that FGFs induce limb formation: FGF protein induces ectopic limb formation; and FGF10-/- mutants lack limbs.
What genes control forelimb vs. hindlimb identity? What experiments demonstrate this?
  • Tetrapod forelimbs and hindlimbs develop in different ways. Thus, there must be differences in gene expression that give rise to these differences in structure. First of all, regions where forelimbs and hindlimbs develop have different sets of Hox genes. Thus, Hos genes must in some way activate different gene expression patterns in forelimbs and hindlimbs.
  • As again in development of ectodermal appendages and the endoderm, it is the mesoderm that confers regional specificity. This is indicated by tissue transplantation experiments. Wing bud (forelimb) mesenchyme was replaced with leg bud (hindlimb) mesoderm. If the mesenchyme controls limb identity, then the forelimb ought be replaced by a hindlimb.
  • As expected, mesoderm (mesenchyme) controls limb identity via regional specificity of induction. This is much like the experiments on epithelial-mesenchymal interactions during ectodermal appendage formation and also in studies of endodermal organ formation. The mesoderm determines the ectodermal appendage or endodermal organ formed.
What roles do Hox genes play in limb development?

Limb Bud Initiation
Hox genes are expressed along the anterio-posterior axis. Hindlimbs and forelimbs appear at different places along the anterio-posterior axis. Furthermore, hindlimbs and forelimbs have different identity. Hox gene expression is responsible for this, as it induces various Tbx genes which the mesodermal mesenchyme conferring regional specificity to the overlying ectoderm.

Limb Bud Patterning

Evidence for existence of a “prepattern”: digits can form in absence of polarizing activity
A final feature I want to mention to you is that the ability of the limb mesenchyme to form segmented structures is instrinsic to the limb bud cells themselves. If we take the cells, diserse them so that we no longer have any gradients of FGF or Shh, and then put them back into an ectodermal limb jacket, a disorganized limb will grow out, but what is striking is that it has recognizable digits composed on individual bones. So the e ability to form a repeated segmental structure is present and doesn’t involve gradients of the patterning molecules we already discussed. This is somewhat reminiscient of the situation with the somites, where we saw the somite formation occurs due to an oscillation of a negative feedback pathway.

Generating Periodic Structures
Reaction-diffusion mechanisms: a form of lateral inhibition established by negative feedback loops.
Observations of this sort have led to the theory that reaction-diffusion mechanisms are involved. These are related to the negative feedback pathway we discussed for somite formation.
The difference is in the reaction diffusion mechanism, the activator protein that gets induced is stable, but it forms a concentration gradient.

Forelimb vs Hindlimb is determined by Tbx = T box (DNA binding domain) transcription factor. Tbx5 determines wing bud and Tbx4 determines leg bud. In the previous slide, we saw that the mesoderm differentiates according to its own developmental program, which was to make leg structures. The positional signals for regional identity must be present in the mesoderm. This implies that there is differential gene expression in wing vs leg.

Whether a hindlimb or forelimb develops is determined very early, by the time a limb bud is evident. The identity will be under the control of Hox genes, since limb identity depends on where on the AP axis the limb bud develops.

Limb identity is determined by the ability of Hox genes to activate the expression of Tbx genes. These are transcription factors. The Tbx genes that control limb identity are Tbx4 and Tbx5.

Now we’ll look at some experiments that demonstrate that Tbx genes control limb identity.
One way to look at this is to induce an ectopic limb to see what type of limb develops, either forelimb or hindlimb, and then to correlate that with the expression of Tbx4 or Tbx5.
In this experiment, a limb bud can be induced in the lateral plate mesoderm by implantation of a bead soaked in FGF. In this case, the bead is implanted in a location where Tbx5 is induced, and the limb develops as a wing.

However, infecting lateral mesoderm with a virus that causes overexpression of Tbx4 will override the normal course of development by altering expression patterns of Tbx genes. Tbx4 is overexpressed in the flank mesoderm by introducing a vector containing a Tbx4 gene. The limb but that now deelops expressed high levels of Tbx4, the gene normally expressed in hindlimb bud. The wing bud now develops into a leg instead of a wing.

We have seen that the PZ controls P-D identity. This must involve differential gene expression. We have already seen that Hox genes control identity along the A/P axis of the embryo and the positioning of the limb buds. They also control identity along the P/D and A/P axis in the limb.

The expression patterns change during development. However, if we look at the expression of Hoxa9 through Hoxa13 in an early limb bud, we can see a colinearity of expression, with the most distal regions expressing the most members of the complex.

Evidence that Hox genes are involved in P/D patterning comes from looking at the phenotypes of mice in which Hox genes are knocked out. For example, were is a normal mouse forelimb. In this mutant, all of the Hox11 genes (hoxa11, d11,etc) are defective. As a result of having no normal Hox11, the foreleg (radius and ulna are almost completely missing).

What we don’t see are homeotic transformations.

Here is an example of a Hox mutation in humans. As we saw in the previous slide, Hoxa13 is expressed in the most distal cells in the limb. In people lacking hoxa13, it is the most distal elements, the digits, that are affected. The types of malformations are complicated and not understood on a molecular level.

What is important is not the type of malformation, but rather the type of structure that is affected by a particular Hox mutation. So, the Hox genes that show the most distally restricted expression affect the most distal limb elements. Hox genes like Hoxd11 that show a more proximal range of expression affect more proximal structures.

Limb developmentComments