Fig 1.
Simplified model of Drosophila network.
A-B: simulated maternal and gap gene profiles. C: simulated pair-rule gene profiles. A’-C’ Summary of interactions used to generate these profiles. Equations and references for the interactions are given in the S1 Text. A generic spatially uniform activator is assumed where needed.
Fig 2.
A Predicted evolutionary pathways from different simulations detailed in Fig 3 (label F1), Fig 4 (label F2), Fig 5 (label Ftz), Fig 7 (label LC). The times shown for the intermediates are only schematic. Gap and eve patterns in three insect species and the inferred last common ancestor (LCA) are indicated. B Summary of homology between Eve modules in different species predicted by our evolutionary simulations.
Fig 3.
Simulated evolutionary pathway (label F1 on Fig 2A) from Drosophila to Anopheles, with salient changes discussed in the main text.
For each transcriptional eve module only the gap genes that regulate it are shown with the same color scheme as Figs 1 and 2A. eve stripe 1 is not shown and the maximum expression of each module is normalized to 1 except when it dips beneath a threshold equivalent to its loss.
Fig 4.
Simulated evolutionary pathway (label F2 on Fig 2A) from Drosophila to Anopheles, following the conventions of Fig 3.
In C the posterior eve 5 stripe is counted as merged with stripe 6.
Fig 5.
Simulated evolutionary pathway (label ftz on Fig 2A) from Drosophila to Anopheles, including ftz.
Conventions of Fig 3 for eve and ftz stripes are used.
Fig 6.
A model for the LCA that imparts stable phase relations among the pair-rule genes and remains consistent with the evolutionary pathway from fly to mosquito.
(A) Schematic of the model showing gap input to eve only. The intensity of repression among the remaining genes (chosen arbitrarily as the primary pair-rule genes in fly) is shown by the line intensity and defines their relative phase. (B) Behavior of the model in A in response to imposed temporal oscillations of Eve, showing phase relationships between different pair-rule genes. Viewed within a cell, one cycle of temporal oscillations would result from the forward shift of the entire Eve pattern by one period. (C,D) If we implement a forward shift of eve by one stripe (left to right panels), by suitably scaling the maternal gradients, then an arbitrary initial arrangement of the three remaining genes is reset to the proper phasing for fly. We show the gap gene configuration for fly in (C) and for mosquito in (D). S3 and S4 Videos show the evolution from the left to right panels respectively for panels (C) and (D). For simplicity only, the model is applied across the entire embryo, though in reality the anterior gap gene input to the primary pair rule genes can remain invariant.
Fig 7.
Simulated evolutionary pathway (label LC on Fig 4) from a presumptive LCA back to Drosophila.
4 steps of evolution are shown (A-D), details are given in the main text. Gap genes profiles are shown as well as activity of different transcriptional eve modules. eve1 is not simulated.