Robustness of cell type identity in Drosophila embryos depleted for bicoid. Max V. Staller¹, Meghan D. Bragdon¹, Zeba B. Wunderlich¹, Norbert Perrimon², Angela H. DePace¹. 1) Department of Systems Biology, Harvard Medical School, Boston, MA; 2) Department of Genetics and Howard Hughes Medical Institute, Harvard Medical School, Boston, MA.

   Developmental gene regulatory networks generate discrete cell types by buffering genetic and environmental variability to produce precise outcomes. It remains unclear how robustness to perturbation emerges from features of network architecture, such as node identities, connection strength, and network topology. To identify the relevant features, we quantify how the Drosophila embryonic segmentation network responds to a severe genetic perturbation. We removed bicoid, a key node in the network, and looked for new cell types as defined by cellular gene expression profiles. We developed a maternal Gal4 short hairpin RNA interference (shRNAi) strategy to deplete bicoid in blastoderm embryos. We quantitatively measured the expression patterns of 12 genes in the anterior-posterior patterning network at cellular resolution by in situ hybridization and two-photon microscopy. Using established methods, we combined data from many embryos into a computationally amenable gene expression atlas, which captures both the direct and indirect effects of bicoid knockdown. To our surprise, removing a key node from the network did not create any new cell types; instead, virtually all cell types in the bicoid-depleted embryo corresponded to cell types in the posterior half of the wild type embryo. Simple models of morphogen-based patterning fail to predict these data; we are currently developing alternative mathematical models of this highly interconnected network to contextualize our results. Analogous to the classical genetic experiments that uncovered the wiring of gene regulatory networks, our quantitative analysis will reveal how network architecture contributes to emergent properties such as canalization and robustness.