(287a) A Facilitated Diffusion Mechanism Establishes the Drosophila Dorsal Gradient

During the first three hours of Drosophila embryogenesis, the transcription factor Dorsal (homologous to NF- κB) patterns the dorsal-ventral axis. Dorsal is present in a nuclear concentration gradient with the highest concentration at the ventral midline and a steady decay to about 40% of the embryoâ??s circumference. The Dorsal gradient is initialized on the ventral side of the embryo via Toll signaling, which phosphorylates the inhibitor protein Cactus (homologous to IκB), marking it for degradation. In the absence of Cactus, Dorsal can enter the nuclei and activate expression of target genes in a concentration-dependent manner. Recent work measuring the dynamics of the Dorsal gradient has shown that, while Dorsal nuclear levels are initially uniform, Dorsal protein diffuses towards the ventral midline, which causes the overall accumulation of a Dorsal gradient peak. To explain this accumulation phenomenon, which seems to act against standard Fickean diffusion, we hypothesize a facilitated diffusion, or "shuttling" mechanism. The Dorsal/Cactus system has each of the features required for facilitated diffusion: (1) Dorsal binds to a â??carrierâ? molecule (Cactus) that protects it from capture; (2) the Dorsal/Cactus complex is diffusible; and (3) the complex is broken in a spatially-dependent manner.

After using a photoactivatable GFP to show that nucleus-to-nucleus diffusion of Dorsal occurs within the embryo, we extended a previous computational model of Dorsal/Cactus interactions. Our model generically predicts that shuttling occurs in the embryo, and that three experimental perturbations would reveal a shuttling-specific phenotype. First, embryos with a half dose of Dorsal develop gradients that are wider and flatter than normal. Second, slowing Dorsal or Cactus diffusion causes the gradient to widen. And third, expanding the spatial domain where Toll signaling causes Cactus degradation results in a widening or even splitting of the Dorsal gradient. We performed each of these model-guided experiments and observed the predicted phenotypes.

Our results suggest that Cactus plays multiple roles in dorsal-ventral axis specification. Besides its predominant role, which is to maintain Dorsal in the cytoplasm, a secondary, but important role is to shuttle Dorsal towards the ventral midline. Given that this mechanism has been found in other, independent systems, we suggest it may be more prevalent than previously thought.