The Bicoid (Bcd) protein gradient in Drosophila serves as a paradigm for gradient formation in textbooks. To explain the generation of the gradient, the ARTS model, which is based on the observation of a bcd mRNA gradient, proposes that the bcd mRNA, localized at the anterior pole at fertilization, migrates along microtubules (MTs) at the cortex to the posterior to form a bcd mRNA gradient which is translated to form a protein gradient. To fulfil the criteria of the ARTS model, an early cortical MT network is thus a prerequisite. We report hitherto undiscovered MT activities in the early embryo important for bcd mRNA transport: (i) an early and omnidirectional MT network exclusively at the anterior cortex of early nuclear cycle embryos showing activity during metaphase and anaphase only, (ii) long MTs up to 50 µm extending into the yolk at blastoderm stage to enable basal-apical transport. The cortical MT network is not anchored to the actin cytoskeleton. The posterior transport of the mRNA via the cortical MT network critically depends on maternally-expressed αTubulin67C and the minus-end motor Ncd. In either mutant, cortical transport of the bcd mRNA does not take place and the mRNA migrates along another yet undisclosed interior MT network, instead. Our data strongly corroborate the ARTS model and explain the occurrence of the bcd mRNA gradient.

The Bicoid (Bcd) protein gradient in Drosophila serves as a paradigm for gradient formation in textbooks. The SDD model (synthesis, diffusion, degradation) was proposed to explain the formation of the gradient. The SDD model states that the bcd mRNA is located at the anterior pole of the embryo at all times and serves a source for translation of the Bicoid protein, coupled with diffusion and uniform degradation throughout the embryo. Recently, the ARTS model (active RNA transport, synthesis) challenged the SDD model. In this model, the mRNA is transported at the cortex along microtubules to form a mRNA gradient which serves as template for the production of Bcd, hence little Bcd movement is involved. To test the validity of the SDD model, we developed a sensitive assay to monitor the movement of Bcd during early nuclear cycles. We observed that Bcd moved along the cortex and not in a broad front towards the posterior as the SDD model would have predicted. We subjected embryos to hypoxia where the mRNA remained strictly located at the tip at all times, while the protein was allowed to move freely, thus conforming to an ideal experimental setup to test the SDD model. Unexpectedly, Bcd still moved along the cortex. Moreover, cortical Bcd movement was sparse, even under longer hypoxic conditions. Hypoxic embryos treated with drugs compromising microtubule and actin function affected Bcd cortical movement and stability. Vinblastine treatment allowed the simulation of an ideal SDD model whereby the protein moved throughout the embryo in a broad front. In unfertilized embryos, the Bcd protein followed the mRNA which itself was transported into the interior of the embryo utilizing a hitherto undiscovered microtubular network. Our data suggest that the Bcd gradient formation is probably more complex than previously anticipated.

No abstract

Supporting information S1 Table. bcd +5+8 phenotypes. Percentages of cuticular phenotypes of 3 h hypoxic and 36 h recovered embryos (left) and bcd +5+8 embryos (right). 3 classes were compared, normal cuticle (blue), mild cuticle phenotype (red) and severe cuticle phenotype (green).