Aaron Smith and Dr. Jeff Barrow, Physiology and Developmental Biology
The apical ectodermal ridge (AER) is a small strip of epithelial tissue that is found at the distal tip of the embryonic vertebrate limb and runs from anterior to posterior. It is known that the AER is necessary to maintain the life of mesenchymal cells (interior non-surface cells) of the limb and is also necessary to maintain other important signaling centers in the limb (citations). The AER is also believed to be responsible for patterning the limb along the proximal to distal axis. However, it is unknown what precise role the AER plays in this and researchers have been attempting to uncover this role over the past forty to fifty years (Dudley et al., 2002).
The Barrow lab has shown that the AER changes shape over time such that it is short and rounded during early limb formation, lengthens and becomes less wide during the middle of limb formation, and becomes long and thin in the final stages of limb development. We have also shown that mice lacking Wnt5a (a secreted molecule) and receptor tyrosine kinase-like orphan receptor 2 (Ror2) have short and fat limbs. Our model of the AER’s involvement in limb patterning is that it secretes Fgf8 which induces the secretion of Wnt5a from nearby limb mesenchymal cells. We hypothesize that the secreted Wnt5a signals through Ror2 to polarize nearby cells so that they will divide and move distally towards the highest concentrations of Wnt5a which are nearby the AER. This concentration gradient that is high near the AER then allows the AER to recruit cells so that the AER’s shape induces the shape of the limb.
Our lab has been able to provide many of the pieces of data to support this model over the last few years, but one piece of data has been very difficult to obtain. We have been attempting to show that a group of cells (or clone) with disabled Ror2 would not grow/migrate towards the AER and that a labeled clone with other gene expression normal would grow towards the AER. Our final and successful attempt to show this was the project I was involved in. We windowed 80-hour chick embryos and sonoporated in constructs that would test the role of Ror2 in the migration and directional division of cells. T
he first step in this process was the generation of the constructs we wanted to use. Initially we wanted to use an IRES GFP to mark transfected cells, and have the plasmid’s expression be tamoxifen inducible. Unfortunately our IRES did not drive expression at a high enough level to easily visualize the transfected cells. To rectify this we decided to cotransfect pCAGGs with a plasmid expressing our genes of interest. The drivers for our genes of interest were a beta actin enhancer and a CMV promoter. This prevented us from utilizing the tamoxifen inducible system since pCAGGs is not tamoxifen inducible so we needed to develop a new way to generate localized clones.
As we practiced injecting the limb buds of the embryos with our constructs/bubble solution prior to sonoporation, we found that sometimes we produced localized clones that met our requirements as far as location, intensity, and size. We wanted to maximize our ability to make these clones and decided that the step of injection was likely to be the largest factor in all the variability we were seeing. Our biggest problem with our preliminary results was that we were leaving behind proximal to distal tracks of transfected cells so we actually began with linear clones. This made it very difficult to prove whether or not the cells were growing linearly. These tracks were created because we were inserting the needle from proximal to distal and our bubble/DNA solution was filling the needle track as we removed it. To avoid these issues we tried several different injection methods. The one that worked the best after practice was to inject from posterior to anterior. It was a much more difficult injection because of the position of the embryo, but we were able to obtain much cleaner results.
We also wanted to improve the intensity and size of the clones because some of the clones were very patchy or were not very bright. We fixed this by using 2ug/uL of pCAGGs in our bubble solution. This maximized our brightness if the injection were successful. Also, we improved the success of our injection by adding India ink for visualization and better removing the amnion and the vitelline membrane prior to injection.
During the course of our experimentation we noted that our construct plasmid seemed to have a cytotoxic effect. Our experimental sonoporations had far less GFP glow than our control sonoporations. To tackle this issue we lowered the dosage of our experimental plasmids so that we would not overload the cell with too much of our dominant negative genes. Lowering our experimental plasmid concentrations to 1ug/uL resulted in good brightness without loss of the dominant negative effect of our genes.
The two dominant negatives we utilized were Ror2 C and Ror2 CRD. C was lacking the intracellular domain and CRD was lacking a cysteine rich region from the extracellular domain. We found that clones expressing Ror2 C did not grow linearly and essentially formed balls of cells as would be expected if there were no directional bias for cell division or migration. As expected, the GFP marked control clones expanded linearly from proximal to distal.
References
- Dudley AT, Ros MA, Tabin CJ. A re-examination of proximodistal patterning during vertebrate limb development. Nature. 2002 Aug 1;418(6897):539-44.