Research in the Brachmann Lab:
Developmental Apoptosis

Drosophila melanogaster is an ideal organism for investigating apoptosis in its full complexity; apoptotic pathways in the fly are closely conserved with mammals and powerful genetic strategies are facilitated by the recent availability of the sequence of the entire fly genome.

As in other organisms, the fly apoptotic process is essential for the removal of damaged cells and is an integral part of development, shaping tissues such as the head, gut, CNS and eye. My interest lies in understanding how cell death during tissue maturation is regulated both spatially, in the larger context of the tissue, and within each cell. What are the signals that tell a cell to die? How are these signals confined to the proper cells? How are they transmitted from outside the cell to inside? How does the receiving cell begin the cell suicide process? How does this process differ in a cell that is induced to die from within by DNA damage?

Our laboratory is investigating these questions by studying the developing Drosophila retina. The Drosophila retina provides a simple model system composed of approximately 750 identically-patterned unit eyes known as ommatidia. Each adult ommatidium contains exactly twenty cells (Figure, right panel): eight photoreceptor neurons (not shown), four cone cells (c, blue), bristles (gray), primary pigment cells (1°, orange) and the interommatidial lattice cells (red). As the pupal eye matures, approximately one-third of the lattice cells are removed by programmed cell death (Figure). The remaining lattice cells stretch to form a precise hexagon around each ommatidium. Similar to, e.g. the mammalian hindbrain, the apoptotic cell is not determined by its lineage but rather by local cell-cell signaling.

      At the time of PCD induction, both dEGFR (Drosophila Epidermal Growth Factor Receptor) and Notch are expressed exclusively in the cells of the lattice (red). Pro-survival signaling to the lattice cells is provided through the TGFa ortholog Spitz, secreted by cone cells (blue), that activates dEGFR and the Ras pathway. In opposition to this, signaling through Notch induces lattice-cell programmed cell death. We have shown that Notch is activated in all lattice cells and that overexpression of Spitz in a lattice cell rescues all neighboring lattice cells. This suggests that normally only a subset of lattice cells respond to a Ras pathway-activating life signal that results in immunity to the Notch-transduced death signal received by all lattice cells. However, no subset of lattice cells is positioned to receive a discreet life signal since every lattice cell shares borders with identical cells prior to PCD initiation.

So then, how are the doomed cells chosen from a group of seemingly identical cells? If all lattice cells are indistinguishable, how are the ill-fated cells selected? Do lattice cells die in a process whereby competing signals determine life or death randomly? If so, how could such a mechanism be precise enough to leave exactly 9 lattice cells and 3 bristle cells around every ommatidium? If not a random process, then how is it regulated?
     
To begin to unravel this mystery, we created a novel method using time-lapse video microscopy and membrane-localized GFP to observe developmental apoptosis as it occurs in the pupal retina (Figure, membrane-bound GFP is black). Our observations definitively revealed that lattice cell death is not random; most death (dying cells are red) occurred along the lateral faces between ommatidia, the region we have termed the "death zone" (circled in pink). These results led us to hypothesize that the "spatial regulator" for death resides subcellularly in the 1° pigment cell to create the "death zone". We are currently investigating.one such localized protein.