RESEARCH

Big picture

Developmental biology seeks to understand how a single cell, the fertilized egg, can give rise to all the complexity of form and function observed in a multicellular organism. One way to address this extraordinarily complicated question is to approach one early and fundamental aspect: in the earliest stages of development, how is head distinguished from tail, and back from front? What are the positional cues that specify the initial asymmetry along these body axes, and how are the structures the structures in between elaborated?

In many organisms, the first indications of these body axes are already visible at the single cell stage. In some cases asymmetry is triggered by fertilization, when sperm entry provides a cue that defines one end of a future axis. In other cases, the unfertilized egg itself contains molecular information, deposited by the mother during oogenesis, that will ultimately specify the poles of the head-tail, or “anterior-posterior,” and back-front, or “dorsal-ventral,” axes. Since we can trace the origin of these basic body axes back to asymmetries at the single cell stage, addressing larger questions about embryonic development begins with an understanding of the mechanisms that generate asymmetry in the egg. As described below, in these issues we can recognize fundamental themes in developmental biology, such as how an inductive signal can trigger cell fate changes or how molecules can adopt specific subcellular localization patterns.

Why the fly?

We study the generation of asymmetry during the development of the fruit fly, Drosophila melanogaster. One of the major advantages of this model organism is the ability to isolate mutations that disrupt a particular process of interest; identification of the affected gene thus reveals a component critical to the proper function of that process. This approach is powerful because it allows the identification of developmentally important genes, even with no prior knowledge of their possible function. Furthermore, the Drosophila genome project continues to provide a wealth of tools – the complete genome sequence, gene expression data, libraries of mapped random mutations, etc. – that allow the increasingly rapid identification of the genes affected in mutant strains and facilitate the study of particular genes of interest.

Generation of asymmetry during Drosophila oogenesis

Figure 1a		  

Our work focuses on the generation of asymmetry during the development of the Drosophila egg. In the ovary, each oocyte develops while interconnected with fifteen sister nurse cells, which function as a source of RNAs and proteins for the developing oocyte and degenerate by the end of oogenesis. Surrounding this cluster of germline cells is an epithelium of somatic follicle cells, which provide yolk and secrete the eggshell (Figure 1a).


Figure 1b		  

Each of these developmental units is called an egg chamber, and the ovary is made up of about sixteen long strings of egg chambers, each functioning as a miniature assembly line for egg production. Notice that the mature egg (Figure 1b) has a complex structure, exhibiting both an anterior-posterior (AP) and a dorsal-ventral (DV) axis. The complex structure of the eggshell implies that the follicle cell epithelium must be regionally specialized in order to secrete an eggshell with the proper shape and surface structures.

How do the cells in this initially uniform epithelium sense their position and acquire the ability to produce the appropriate eggshell features? The embryo develops with its axes in register with those of the egg, instructed by localized molecular information incorporated into the egg as it develops.

Polarity along the DV axis

Our work focuses on how asymmetry is initiated and elaborated along the DV axis. This axis is defined by a growth factor, called Gurken, produced by the oocyte. The key to generating asymmetry is that Gurken is only found on the dorsal side of the oocyte, where it activates a receptor tyrosine kinase (the Drosophila homolog of the vertebrate EGF receptor [Egfr]) in the adjacent follicle cells (Figure 2).


Figure 2		 Receptor activation triggers a signal transduction pathway that directs follicle cell fate and thus distinguishes dorsal cells, which receive the signal, from ventral cells, which do not. This signaling process, which is refined over time by downstream feedback mechanisms, ultimately determines the structure of the eggshell along its DV dimension.
In mutants where either the signal or receptor is absent, no follicle cells are signaled to produce dorsal-appropriate eggshell structures and instead only ventral-like eggshell is secreted. What does the eggshell have to do with the embryo? While the dorsal follicle cells secrete more elaborate eggshell structures, the ventral follicle cells produce an unknown secreted molecule that acts as the spatial cue that defines the DV axis future embryo. Thus the patterning events initiated by Gurken ultimately govern both eggshell and embryonic development.

RESEARCH OVERVIEW


Figure 3

A mutation affecting eggshell patterning

One of our research goals is to understand the series of events that are triggered by the initial Gurken signal and instruct the dorsal follicle cells to secrete the dorsal side of the eggshell, including its prominent eggshell appendages. To approach this question, we are studying a gene called capicua. Females lacking capicua function produce eggs with dramatic eggshell defects: the dorsal appendages are very broad and shifted toward the ventral side, with a wider space between them (Figure 3). This phenotype reflects an expansion of the dorsal side of the eggshell and therefore suggests that more follicle cells have adopted a “dorsal” fate.

Increased dorsal follicle cell fate determination suggests an increase in Gurken signaling; indeed, other mutants with similar “dorsalized” phenotypes exhibit an expanded distribution of Gurken within the oocyte. Surprisingly, however, Gurken is normal in capicua mutant ovaries, suggesting that loss of capicua function affects follicle cell fate at a later step. Since mutations in capicua appear to disrupt follicle cell patterning in a previously undescribed manner, the study of capicua function may yield novel insights into how the pattern of follicle cell fates is determined.

Establishment of DV polarity in the germline

To complement our studies of DV patterning in the follicle cell epithelium, we are also investigating the mechanisms that generate asymmetric Gurken localization in the oocyte. The localization of the Gurken protein reflects the localization of its mRNA; in general, subcellular RNA localization requires cis elements within the message as well as subcellular machinery to carry out proper targeting. For the gurken mRNA, regions important for localization have been identified, but how they function is not understood. While some recent work has demonstrated that injected gurken transcripts can be transported dorsally via the oocyte microtubule network, it is unclear whether this mechanism is relevant in vivo. As a first step to understanding how Gurken is localized, we are investigating the source of the gurken mRNA. Is it produced in the nurse cells and transported to the oocyte, like many other oocyte components? Or is it produced in the oocyte itself? Understanding the source of gurken production will allow better evaluation of potential localization mechanisms.

The follicle cell epithelium as a model for wound healing

Figure 4a

Figure 4b

A final project explores the interactions between follicle cells, rather than their role in DV patterning. This work was initiated when we identified a mutation with a striking effect on the organization of the follicle cell epithelium: when we generate genetically mosaic epithelia composed of both mutant and wild type cells, we find that groups of mutant cells are separated from their neighbors by a smooth boundary. We first noticed this phenotype in the pattern of the follicle cell imprints on the eggshell surface (Figure 4), then looked directly at the follicle cells themselves during oogenesis. Interestingly, in addition to a smooth boundary, levels of filamentous actin are increased at the interface between wild type and mutant cells.

These phenotypes resemble the cellular response to small wounds in epithelial monolayers in both vertebrate model systems and Drosophila embryos, where actin accumulates at the edge of each cell facing the wound and forms a “contractile ring” that constricts to close the wound and preserve the integrity of the epithelium. In vertebrate cultured cell systems, cells undergoing programmed cell death are extruded from an epithelial monolayer by a similar mechanism, again maintaining the barrier function of the epithelium.Are our mutant cells provoking a wound healing response in their neighbors? This general question has two components. First, what is the nature of the defect in our mutant cells? Are they dying and inducing a wound healing response in their neighbors? Though they are able to secrete eggshell and seem to be intact through oogenesis, we are testing for early signs of programmed cell death. We are also working to identify the gene affected by this mutation, which may reveal the underlying cellular defect. Whatever the defect, how do neighboring cells detect and respond to this change? Though recent work has identified some of the molecules that are required for contractile ring assembly, constriction and healing, little is known in Drosophila or other systems about how a cell or group of cells can signal this type of response in adjacent cells. We hope that this work may ultimately allow us to use the genetics of this model organism to address these cell biological questions.