Generation and timing of graded responses to morphogen gradients

ABSTRACT Morphogen gradients are known to subdivide a naive cell field into distinct zones of gene expression. Here, we examine whether morphogens can also induce a graded response within such domains. To this end, we explore the role of the Dorsal protein nuclear gradient along the dorsoventral axis in defining the graded pattern of actomyosin constriction that initiates gastrulation in early Drosophila embryos. Two complementary mechanisms for graded accumulation of mRNAs of crucial zygotic Dorsal target genes were identified. First, activation of target-gene expression expands over time from the ventral-most region of high nuclear Dorsal to lateral regions, where the levels are lower, as a result of a Dorsal-dependent activation probability of transcription sites. Thus, sites that are activated earlier will exhibit more mRNA accumulation. Second, once the sites are activated, the rate of RNA Polymerase II loading is also dependent on Dorsal levels. Morphological restrictions require that translation of the graded mRNA be delayed until completion of embryonic cell formation. Such timing is achieved by large introns, which provide a delay in production of the mature mRNAs. Spatio-temporal regulation of key zygotic genes therefore shapes the pattern of gastrulation.


Fig. S2. Binding of Dorsal and Zelda
CHIP seq data from (Sun et al., 2015) was plotted. Right-pointing arrows mark transcription start sites and the extent of the transcribed regions are shown by bold black lines. A) Binding of Dorsal to the twi regulatory region. A') Binding of Zelda to the twi regulatory region. Prominent binding of both is detected immediately upstream to the TS. B,B') Binding of Dorsal and Zelda to the T48 regulatory the region. Prominent binding is detected upstream to the TS and in the first intron. C.C') Binding of Dorsal and Zelda to the mist regulatory region. Prominent binding is detected upstream to the TS and for Zld also within the transcribed region.

Fig. S3. Pol II Pausing
Chip seq data of Pol II from (Saunders et al., 2013) was plotted. Right-pointing arrows mark transcription start sites and the extent of the transcribed regions are shown by bold black lines. Prominent stalling was observed. A) twi. B) T48. C) mist.

Fig. S4. Modelling Dorsal-dependent mRNA accumulation
Based on experimental results, Dorsal was assumed to determine both the promoter activation probability and Pol II loading rate of the T48 gene. Mature transcripts are assumed to be stable and accumulate during the course of the simulation (~15 minutes). Differential equations used to simulate the dynamical system-Activation probability (pactivation) and Pol II loading rate (rpol2) depend linearly on Dorsal level and thus on the position along the DV axis. The dynamics of the activated promoter fraction (Factivated) and the mRNA level (mrna) is simulated for 15 minutes after model initiation (no mRNA or activated promoters at time 0). Model parameters were either arbitrary (rpol2 at maximal Dorsal levels) or based on experimental results (pactivation at maximal Dorsal=0.07 min -1 ) (Ambrosi et al., 2014).

Fig. S5. Uniform activation of mist in Toll DLRR embryos
Uniform expression of the constitutive Toll DLRR construct drives activation of the pathway and nuclear targeting of Dorsal along the entire embryo circumference. Induction of twi (green) and mist (red) TSs responds accordingly. In a younger embryo (A-B') one or two twi TSs are observed in most nuclei, while only a single or no mist TSs are observed in many nuclei. In an older embryo (C-D') the majority of nuclei exhibit two TSs for both genes. Scale bar 10 µm.

Fig. S6. Determination of embryo age at NC 14
The reported apical-basal lengths of DAPI-stained nuclei were calculated as the average length of 20 nuclei on both sides of the pole cells (PC) . This value was used to estimate the embryo age in NC 14 at 20 0 c, according to (Lecuit and Wieschaus, 2000). The embryos shown correspond to the panels in Figure 7. A) 8µm, B) 5.4 µm, C) 7.2 µm, D) 12.5 µm. Scale bar 10 µm. Development: doi:10.1242/dev.199991: Supplementary information

Movie 1. Dynamics of Myosin II recruitment
The dynamic localization pattern of Myosin II during early NC 14, monitored using a GFP "protein-trap" in the Myosin II heavy-chain (zipper) gene locus, as visualized in cross section by Lightsheet microscopy. During the process of cellularization, Myosin II is initially recruited to the basal aspect of the future cell membranes as they extend from the surface. Upon completion of cellularization, Myosin II disappears form the basal position and is recruited to the apical side in the ventral-most cells, where it drives furrow formation. The graded pattern of Myosin II recruitment within the ventral domain shapes the pattern of cell invagination. The movie was obtained at a rate of 1 frame/min.