Dorsal Ventral Polarity in Drosophila


Team Members: Hector Lopez & Raj Daniels

References & reference Images:

Animations
*click on animation to view*
Animation 1
Anitmation 2

Comments :
Dorsal-ventral polarity in Drosophila is accomplished by the gradient of a transcription factor known as Dorsal. The specification of cell types (i.e. dorsal or ventral) along the dorsal-ventral axis is established by the interactions of many cytoplasmic materials within the single multinucleated cell, which will be described subsequently. The most important step in the establishment of dorsal-ventral polarity is the translocation of the dorsal protein from the cytoplasm to the ventral cell nuclei. Dorsal protein plays a central role and it acts on cell nuclei to specify different regions of the embryo, and their respective cell fates which are all regulated by concentration gradients.

In specification of the dorsal-ventral axis, 11 maternal effect genes have been identified. The products of these genes are required for proper development, and the absence of just one product results in the lack of ventral structures (Anderson). It is interesting to note that genes in Drosophila are usually named after their mutant phenotypes. For example, the dorsal gene is needed for the differentiation of the ventral cells. The lack of the dorsal gene causes the ventral cells to become dorsalized. This also applies to other genes in this process such as snake and easter. The absence of snake causes a "snake" shaped larvae to develop, while easter results in larvae with an easter egg appearance (Anderson).

To start with, the oocyte nucleus is originally located at the posterior end away from the ovarian nurse cells. However, it will eventually move to an anterior dorsal position and synthesize the product of gurken. The Gurken protein product can only diffuse a short distance and thus will only reach those follicle cells closest to the nucleus, which are located on the dorsal side. During mid-oogenesis, the Gurken protein is received by Torpedo receptors, which are located on the follicle cell membrane, after leaving the oocyte and crossing the perivitelline space. This event activates the Torpedo receptor protein, inhibiting the expression of pipe. Consequently, the Pipe protein is made only in the follicle cells on the ventral end, where the Torpedo receptors fail to receive the gurken message and are rendered inactive. The significance of the process mentioned above is that the fate of the follicle cells surrounding the oocyte has been determined. Depending on their location (dorsal or ventral), these cells will then determine which nuclei of the embryo receive the dorsal protein.

Approximately 90 minutes after fertilization, the Dorsal protein is synthesized from the RNA transcript provided by the maternal Dorsal gene (Roth et al.). While this protein can be found throughout the syncitial blastoderm of the early embryo, it is translocated only into ventral nuclei. In ventral follicle cells, the Pipe protein modifies an unknown "X" factor, allowing it to be secreted into the perivitelline space. At first it was thought that Pipe modified Nudel protein, which has been found to have regions for GAG addition that Pipe could act upon. However, research has shown that Nudel function does not require that it be expressed in the ventral follicle cells. This has led to the belief that Pipe acts in the Golgi on GAGs associated with an X-factor, which is likely to be a proteoglycan (Sen et al)

Once the modified X-factor enters the perivitelline space, it is free to interact with the Nudel protein to initiate a cascade of activating events that cleave three serine proteases. First, the Nudel-X factor complex causes Gastrulation defective protein to be split. The resulting activated Gd protein is now able to move on and cleave the Snake protein. The activated snake protein cleaves the Easter protein, which will then go on to cleave the Spatzle protein. The activated Spatzle protein is then able to bind to Toll receptors located on the membrane of the early embryo. Although the Toll protein is evenly distributed, it becomes activated only by binding the Spatzle protein, which is only active in the ventral perivitelline space.

The activation of the Toll protein is critical to the establishment of the gradient of the Dorsal protein in the ventral nuclei. When Toll is activated, it is able to activate the Pelle protein kinase. Associated with Pelle is the Tube protein, which likely causes Pelle to be brought to the cell membrane for activation. Activated Pelle is then able to phosphorylate the Cactus protein, which blocks Dorsal protein from entering the ventral nuclei. If Cactus is not degraded, the Dorsal signal will remain in the cytoplasm. Once Cactus has been removed, the Dorsal protein is able to freely move into the ventral nuclei.

It is important to note that this sequential cascade of events creates a gradient of Spatzle protein that is greatest in the ventral regions of the early embryo, which in turn causes the gradient of Dorsal protein signal to be highest in the most ventral cell nuclei. The nuclei of the early embryo become cellularized at approximately the 14th division cycle. The Dorsal protein is translocated to the appropriate nuclei before this happens. Large amounts of Dorsal protein instruct the nucleus to make a mesodermal cell, while lesser amounts cause the cells to become glial or ectodermal tissue.

The Dorsal protein specifies these cells to express a mesodermal phenotype by activating twist, snail, and rhomboid. These genes are transcribed only in nuclei that have received high concentrations of Dorsal protein, since their enhancers do not bind Dorsal with a very high affinity. In addition, the Dorsal protein determines the mesoderm directly by inhibiting the dorsalizing genes zerknullt and decapentaplegic. Finally, the dorsal-ventral axis is subdivided into mesoderm, mesectoderm, neurogenic ectoderm, epidermis, and amnioserosa in the following way. Mesodermal precursors express twist and snail, but not zerknullt and decapentaplegic. Precursors of dorsal epidermis and amnioserosa express zerknullt and decapentaplegic, but not snail or twist. Mesectoderm (glial) precursors express only snail, while lateral neural ectoderm lateral neural ectoderm precursors do not express any of the four genes.

There are many shortcomings to the diagram in Gilbert as well as our animations. For example, only one nucleus is shown to undergo all of the processes described above. This is obviously not the case, because then the nucleus would be either dorsal or ventral and that would be the end of it. The fate of the follicle cells is, however, determined when only one nucleus is present in the oocyte. But, after the follicle cells have been specified as either dorsal or ventral, fertilization occurs, eventually creating a syncitium (as centrolecital embryos do). Interestingly enough, how does the sperm fertilize the egg surrounded by follicle cells and an "egg shell?" In this way, it might have been helpful to animate the fertilization step. Soon after fertilization, the establishment of the gradient of the Dorsal protein occurs while the embryo is in the multinucleated stage. Therefore, while the diagrams in the book and our animations are zoomed in depictions, it should be noted that the translocation of Dorsal occurs in many nuclei, not just one. Moreover, Gilbert should have labled the cell whose nucleus received Dorsal as an embryo, not an oocyte. Finally, the nuclei become cellularized after the Dorsal gradient has been created. Otherwise, it would be difficult for Dorsal to enter the nuclei with a cell membrane surrounding it.