Dynamics of Spaetzle morphogen shuttling in the Drosophila embryo shapes pattern

Establishment of morphogen gradients in the early Drosophila embryo is challenged by a diffusible extracellular milieu, and rapid nuclear divisions that occur at the same time. To understand how a sharp gradient is formed within this dynamic environment, we followed the generation of graded nuclear Dorsal (Dl) protein, the hallmark of pattern formation along the dorso-ventral axis, in live embryos. We show that a sharp gradient is formed through extracellular, diffusion-based morphogen shuttling that progresses through several nuclear divisions. Perturbed shuttling in wntD mutant embryos results in a flat activation peak and aberrant gastrulation. Re-entry of Dl into the nuclei at each cycle refines the signaling output, by guiding graded accumulation of the T48 transcript that drives patterned gastrulation. We conclude that diffusion-based ligand shuttling, coupled with dynamic readout, establishes a refined pattern within the diffusible environment of early embryos.


26
The crude onset and subsequent refinement of spatial information shapes the future body 27 pattern of embryos. Morphogens, key instructive elements in this context, are secreted 28 signaling molecules that induce cells to adapt different fates depending on their concentration. 29 Establishing a morphogen gradient over a field of naïve cells patterns the cell layer into distinct 30 domains of gene expression (Green and Sharpe, 2015;Wolpert, 1971). Different strategies to 31 guide morphogen distribution have been identified. A common option is to produce the 32 morphogen in a restricted group of cells, giving rise to its graded distribution in the 33 surrounding cells (Lecuit et al., 1996;Nellen et al., 1996). Notably, in this scenario, the 34 morphogen-producing cells are distinct from the responding cells. 35 An alternative strategy of morphogen distribution is applicable to situations where the 36 morphogen is broadly expressed, and the gradient is generated within the field of expressing 37 cells, which also respond to the morphogen. This scenario is applicable to early embryos, 38 where broad transcriptional domains have been established, but have not yet given rise to the 39 determination of sufficiently restricted groups of cells, which could provide a local morphogen 40 source. In such situations, restricting morphogen signaling to a narrow domain becomes a 41 challenge, as diffusion tends to spread, rather than restrict ligand distribution. 42 Studies in several systems identified the Shuttling mechanism as a robust solution to this 43 challenge (Shilo et al., 2013). Here, a morphogen gradient is established not merely by its 44 diffusion away from the production source, but through an effective translocation of the 45 morphogen into the center of the field. This translocation, which is purely diffusion driven, is 46 mediated by a proximally-produced inhibitor. The resulting gradient is sharp and robust, 47 displaying limited sensitivity to gene dosages or reaction rate constants. triggers Dorsal (Dl) translocation into the syncytial nuclei ( Figure 1A). The processed form of 98 the Spz ligand therefore functions as the morphogen at this stage. 99 We previously showed that graded active Spz distribution is established by a shuttling 100 mechanism. In this case, shuttling is implemented in a self-organized manner through a 101 complex interplay between the active ligand and its pro-domain, which can accommodate Entry of Dl into the nuclei can be followed in single live embryos carrying a Dl-GFP fused 106 protein (DeLotto et al., 2007). In this work, we use Light Sheet fluorescence microscopy for 107 live imaging of Dl-GFP nuclear localization during the final nuclear division cycles of the 108 syncytial Drosophila embryo. The resulting dynamics shows the two signatures of ligand 109 shuttling: a transient increase in signaling in the lateral regions, which is then reduced so as to 110 preferentially increase signaling at the ventral midline, and the resolution of two lateral peaks 111 . CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint . http://dx.doi.org/10.1101/505925 doi: bioRxiv preprint first posted online Dec. 26, 2018; to a single central peak. We find that ligand shuttling is an ongoing process, which repeats 112 itself following each nuclear division. During the beginning of nuclear cycle (NC) 14, the 113 resulting dynamics of nuclear re-entry of Dl allows to further refine the resulting spatial 114 pattern, by triggering different temporal onsets of zygotic target gene expression in closely 115 positioned nuclei, thereby leading to a functionally significant graded accumulation of target 116 gene transcripts. In wntD mutant embryos, the Dl peak becomes flattened, and leads to an 117 abnormal increase in the number of cells simultaneously undergoing the initial step of 118 gastrulation, underscoring the significance of timely and properly shaped gradient formation. 119 Thus, diffusion-based ligand shuttling, coupled with a dynamic readout, establishes a 120 refined pattern within the environment of early embryos. V gradient of nuclear Dl-GFP already at NC 12. This gradient was further refined and 128 elaborated during the next two cycles. To enable quantitative analysis of Dl-GPF nuclear 129 dynamics, we used an area preserving transformation to project the 3D images onto a 2D sheet. 130 We restricted our analysis to a region surrounding the A-P midline, where distortion due to 2D 131 projection is negligible ( Figure 1C, SI: movie 2) (Heemskerk and Streichan, 2015). Next, we 132 automatically segmented the nuclei and averaged the nuclear Dl-GFP signal in nuclei 133 occupying a similar D-V axis position. 134 Our measurements defined the quantitative, spatio-temporal dynamics of Dl-GFP at a 1-2 135 minute time resolution ( Figure 1C-D, SI: movie 3, Methods). This dynamics results from the 136 extracellular active Spz gradient. However, inferring the profile of this extracellular gradient 137 from Dl-GFP dynamics is confounded by the fact that Dl-nuclear accumulation is established 138 anew at every nuclear cycle, since Dl exits the nucleus at mitosis upon nuclear envelope 139 breakdown. We therefore needed a framework to suitably infer properties of the extracellular 140 active Spz gradient, and critically distinguish between models of gradient formation. 141 Toll signaling, at each given position along the D-V gradient, triggers the level of nuclear Dl 142 and the rate by which this level increases. Thus, at the beginning of each division cycle, 143 following re-establishment of the nuclear envelope, nuclear Dl levels increase at a rate that is 144 proportional to the level of nearby Toll signaling. Conversely, at longer times, nuclear Dl 145 levels approach a steady state, and are proportional to the extracellular Toll signaling. We 146 therefore plotted the dynamics of both parameters, nuclear Dl levels and its temporal change 147 during the onset of NC 14. Notably, we observe that this qualitative dynamics differs, 148 depending on the spatial position of nuclei along the D-V axis. In the ventral-most regions they 149 increased monotonically. In contrast, in lateral domains nuclear Dl displayed an overshoot, 150 initially increasing but then starting to decrease (Figure 2A-D, Figure S1). Clearly, such a 151 decrease in nuclear Dl is only possible if Toll signaling at this position decreases as well. 152 Therefore, the data indicates that the external Spz gradient continues to evolve through the 153 early part of NC 14, showing a distinct position-dependent, non-monotonic temporal signature. 154 To more rigorously infer dynamic properties of the external gradient from the measured pattern 155 of nuclear Dl, we used computer simulations, modeling Dl-nuclear entry while assuming 156 different temporal patterns of Toll signaling ( Figure 2E-L). Specifically, we searched for a 157 qualitative signature that distinguishes between three scenarios: (1) constant Toll signaling; (2) 158 Toll signaling that is changing (increasing) monotonically in time, as expected in naïve 159 gradient-forming models; and (3) a non-monotonic increase in Toll signaling, the signature 160 found in lateral regions of gradients formed by the shuttling mechanism. Our simulations have 161 shown that these scenarios are best distinguished by comparing the temporal changes in 162 nuclear Dl (d(Dl)/dt) with the levels of nuclear Dl. In the first two cases -constant or 163 monotonically increasing Toll activity -the relation between these two parameters is 164 invariably linear or concave ( Figure 2E-L). In contrast, in the presence of non-monotonic 165 shuttling-based dynamics, a convex relation is obtained, with a pronounced negative temporal 166 derivative at the lateral regions, where nuclear Dl levels are low ( Figure 2M-P). 167 The measured data is not consistent with the dynamic defined by constant, or monotonically 168 increasing Toll signaling. Rather, it shows a clear signature of non-monotonic, shuttling-like 169 dynamics. Extending our simulations to include the full shuttling model that establishes the 170 active Spz gradient combined with Dl nuclear transport (See SI), confirmed that this model is 171 fully capable of simulating the experimentally observed dynamics, including the non-172 monotonic, overshoot dynamics at the lateral regions. 173 An additional notable property of Dl-nuclear entry dynamics was the initial formation, at every 174 nuclear cycle, of two ventro-lateral signaling peaks, that eventually converge to a single ventral 175 peak ( Figures 1D, 2A, Figure S1). Thus, by 10-15 minutes into NC 14, when the major target 176 genes for Dl are induced, the initial two-peak gradient has refined to a single sharp peak. The 177 initial two-peak pattern provides another unique signature of shuttling-like dynamics. It is 178 expected under certain shuttling parameters, when the mean distance traveled by the shuttling 179 complex before it is cleaved, is much smaller than the distance to the ventral-most site. In this 180 case, ligand will initially accumulate at lateral regions, followed by gradual ventral In conclusion, the dynamic behavior of Dl-GFP supports a continuous process of extracellular 187 Spz shuttling, displaying two of its defining signatures: non-monotonic dynamics of nuclear Dl 188 entry in lateral regions, and the transient formation of two-peak gradient. 189

Altered shuttling dynamics in wntD mutants affects gastrulation 190
Dl-nuclear localization dynamics can be used for refined analysis of informative mutant 191 phenotypes. We applied this approach to study WntD, an inhibitor of Toll signaling which 192 provides a negative feedback that buffers the D-V patterning gradient against fluctuations 193 (Rahimi et al., 2016). wntD, a target of the Toll pathway, is transcribed locally at the posterior 194 . CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.  WntD levels would therefore lead to redistribution of the ligand to more ventral regions, 207 impacting not only on the strength of Toll signaling, but also on its sharpness (see Figure 3C To follow the morphological consequences of a wider peak distribution of nuclear Dl 216 distribution, we monitored wntD mutant embryos for an extended period of NC 14, observing 217 the processes of gastrulation and ventral furrow formation. We defined the edges of the 218 furrowing domain by marking the two lateral-most nuclei that alter their orientation upon 219 gastrulation. Working backwards to an earlier phase of NC 14, when the nuclei are still in a 220 monolayer, we can accurately count the number of nuclei between these edges. In contrast to 221 gastrulating wt embryos where the initial invagination is observed in ~9 cells, in wntD mutants 222 a broader front of up to 15 cells invaginated at the same time ( Figure 4). Thus, the shape of the 223 . CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint . http://dx.doi.org/10.1101/505925 doi: bioRxiv preprint first posted online Dec. 26, 2018; Dl-activation gradient is essential for normal patterning and gastrulation. When the final 224 gradient peak is not sharp, a larger cohort of ventral cells takes part in furrow formation. 225

Timing of wntD transcription 226
wntD mutants display perturbed Dl dynamics already at NC 13 ( Figures 3H, S3) suggesting 227 that WntD normally exerts its modulating effects at this early stage. This implies that zygotic 228 expression of wntD, its translation, secretion to the peri-vitelline fluid and diffusion of the 229 protein, have commenced by then. To examine that this is possible, we applied our Light 230 Sheet-based visualization setup as a more sensitive assay for defining the onset and temporal 231 dynamics of wntD expression. 232 . CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. reporter, we find that signal intensity in transcribing nuclei is similar regardless of their 259 position along the D-V axis, suggesting that once T48 transcription is initiated, it progresses at 260 a constant rate in all nuclei ( Figure 6C). If zygotic expression of Dl-target genes depends not 261 only on the final, steady-state level of nuclear Dl, but also on the dynamic profile of its 262 accumulation, the signaling output could be further sharpened. A dynamic phase of Dl nuclear 263 entry takes place during the initial 20 minutes of NC 14 ( Figure 1D), the nuclear cycle 264 associated with a major onset of zygotic gene expression. A consequence of these dynamics is 265 that ventral-most nuclei will reach the threshold for expression of a given zygotic gene earlier 266 than more lateral ones. These ventral nuclei will begin to express the gene earlier, therefore 267 expressing it for a longer period than more lateral nuclei, and could thus accumulate more 268

transcripts. 269
To examine the consequences of the graded onset of T48 transcription on mRNA 270 accumulation, we carried out quantitative single-molecule FISH using T48 probes. utilization of a "self-organized shuttling" mechanism in this context, it was imperative to 303 visualize the actual dynamics of the process. 304 We were able to infer the dynamics of the extracellular Spz gradient by following the kinetics 305 of Dl-GFP nuclear accumulation in individual live embryos during the final syncytial nuclear 306 division cycles and the early phase of NC 14. Nuclear levels of Dl are not a direct readout of 307 the extracellular gradient, since accumulation of Dl in the nuclei is re-initiated at the onset of 308 every nuclear cycle. Nevertheless, it is possible to infer key features of the extracellular Spz 309 . CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint . http://dx.doi.org/10.1101/505925 doi: bioRxiv preprint first posted online Dec. 26, 2018; gradient from this dynamic behavior. Using this approach we identified clear hallmarks of 310 ligand shuttling, most notably the lateral overshoot and the presence of two lateral peaks which 311 converge to a central ventral peak. This convergence takes place within a timeframe of 312 minutes, and repeats at every nuclear cycle. Since new protein molecules of the extracellular 313 components are continually translated, the ongoing activity of the shuttling process is vital. 314 Therefore, shuttling is important not only for generating the gradient, but also for maintaining 315 it, in the face of rapid diffusion and mixing within the peri-vitelline fluid. Importantly, by ~10-316 15 minutes into NC 14, when the robust induction of transcription of the cardinal zygotic Dl-317 target genes twt and sna ensues, the nuclear gradient of Dl is sharp and a single ventral peak is 318 resolved. 319 The role of WntD in shaping and buffering the Spz gradient 320 Having described the dynamics of Dl-nuclear entry and gradient formation, we were in a 321 position to use our experimental approach in order to examine regulatory processes affecting 322 (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint . http://dx.doi.org/10.1101/505925 doi: bioRxiv preprint first posted online Dec. 26, 2018; autonomous, and will actually change the shape of the gradient, making it sharper. The 338 observed dynamics of Dl-GPF in wntD mutant embryos indeed confirms this prediction. 339 The shuttling process is driven by competition between the inhibitory Spz pro-domain and the 340 Toll receptor for binding free, active Spz. Binding to the pro-domain is favored in the lateral 341 part of the embryo, where its concentration is higher, while in more ventral regions binding to 342 Toll takes over. Since WntD impinges on the extracellular properties of the Toll receptor, the 343 active ligand is deposited in more ventral regions, where the concentration of the pro-domain is 344 lower. Thus, WntD does not simply reduce the overall profile of Toll activation, but actually 345 re-directs the ligand from the lateral regions to the ventral domain. We have previously shown 346 that accumulation of excess ligand in the peak by shuttling is an effective mechanism to buffer 347 noise. Since activation in this region is already maximal, the excess ligand will not alter the 348 resulting cell fates (Barkai and Shilo, 2009). 349 The rapid timing of processes in the early embryo and the short duration of interphases 350 between nuclear divisions raises the question of whether it is actually possible to produce 351 sufficient levels of WntD that will drive the morphogen profile to the desired equilibrium. . CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint . http://dx.doi.org/10.1101/505925 doi: bioRxiv preprint first posted online Dec. 26, 2018; These genes are triggered at NC 14 after the Dl gradient is stabilized and a distinct activation 365 peak generated. 366 Are there zygotic target genes that respond to the dynamics of Dl nuclear targeting, before it 367 stabilizes? This appears to be the case for T48, which encodes a transmembrane protein that strict dependence on the timing of transcription initiation provides another mechanism to 379 generate differences between adjacent nuclei along the D-V axis. 380 In conclusion, this work has utilized live imaging of Toll pathway activation, to identify and 381 characterize the hallmarks of ligand shuttling (Figure 7). This process is rapid and takes place 382 continuously throughout the final nuclear division cycles, to generate and maintain a sharp 383 activation gradient in the diffusible environment of the peri-vitelline fluid. WntD impinges on 384 Spz shuttling, and is responsible not only for buffering variability between embryos, but also 385 for generating a sharp activation peak. This peak is utilized to induce a graded expression of a 386 zygotic target gene that is essential for executing processes that drive gastrulation. Thus, 387 diffusion-based ligand shuttling, coupled with a dynamic readout, establishes a refined pattern 388 within the environment of early embryos. CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. To enable quantitative analysis of the nuclear Dl gradient, we projected the 3D scans of the 451 embryo from the light-sheet microscope, into a 2D flat surface, for every time point imaged. 452 This was possible, since all the nuclei are arranged on the surface of the embryo, whose shape 453 resembles an ellipsoid. This ellipsoid can be projected into a 2D surface, which contains all the 454 nuclei and therefore the entire nuclear Dl gradient. To this end, we used an area preserving 455 transformation with minimal distortion far from the Anterior and Posterior poles, implemented 456 by the IMSANE tissue cartography tool (Heemskerk and Streichan, 2015). IMSANE was used 457 with the following specifications: Planar Illastik surface detector and cylinder chart type. 458 Surface detection was performed on the last time point for each embryo, and the detected 459 surface was then used to project all earlier time points. Since the embryo is, to a good 460 approximation, a cylinder apart from the anterior and posterior poles, embryo circumference 461 was defined as the largest circumference of the ellipsoid fitted to the embryo surface by 462

Nuclei segmentation 464
The nuclei were detected separately for each time point, using the following segmentation 465 method: 466 1. Automated local thresholding of the image in order to create a binary mask. Done 467 in ImageJ using the Bernsen algorithm with a contrast threshold of 15. 468 2. The resulting binary mask underwent further refinement to segment the nuclei 469 using MATLAB image analysis filters: 470 a. All connected objects in the mask large enough to be nuclei (over 50 471 pixels in size) were located and classified into 3 size groups: small 50-150 472 pixels, medium 150-600 pixels and large 600 pixels and over. 473 b. Each size group underwent erosion using imerode and then dilation using 474 imdilate with appropriate filter sizes for each group. 475 . CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint . http://dx.doi.org/10.1101/505925 doi: bioRxiv preprint first posted online Dec. 26, 2018; c. Objects belonging to the large group underwent another round of erosion 476 and dilation with same filter size as first round. 477 3. The resulting objects in the binary mask were filtered by size to exclude objects 478 too small (under 50 pixels) or too large (over 3000 pixels) to be a single nucleus.

Dorsal gradient as a function of time, at specific locations along the D-V axis 500
For the calculation of Dl-nuclear intensity over time at a specific location ̃/ , we used 501 ℎ ( / , ) at that location: ℎ (̃/ , ). Background subtracted values were 502 calculated separately for each NC, by subtracting the minimal intensity observed in a nucleus 503 for that NC. ℎ (̃/ , ) was then smoothened in time using the MATLAB smooth 504 . CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint . http://dx.doi.org/10.1101/505925 doi: bioRxiv preprint first posted online Dec. 26, 2018; function with the loess method and a smoothing coefficient of 0.5. It was then fitted with a 505 smoothing spline, resulting in the function ̃/ ( ). The temporal derivative of nuclear 506 Dorsal, at a specific location-̃/ was calculated by applying a third order finite differences 507 formula to ̃/ ( ) and then smoothing using smooth with a smoothing coefficient of 0.6 508 and fitting a smoothing spline. 509 CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.

Measuring peak sharpness
The copyright holder for this preprint . http://dx.doi.org/10.1101/505925 doi: bioRxiv preprint first posted online Dec. 26, 2018;   CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.

Figure Legends
The copyright holder for this preprint . http://dx.doi.org/10.1101/505925 doi: bioRxiv preprint first posted online Dec. 26, 2018; CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.  CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.  CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint . http://dx.doi.org/10.1101/505925 doi: bioRxiv preprint first posted online Dec. 26, 2018; . CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint . http://dx.doi.org/10.1101/505925 doi: bioRxiv preprint first posted online Dec. 26, 2018; CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.  . CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.

779
Movie 1 -A Light Sheet time-lapse movie following the dynamics of endogenously expressed Dl-GFP in the entire embryo.

780
The nuclear Dl gradient can be seen in nuclei at the ventral side (bottom) already at NC 12, it is lost during nuclear divisions 781 and is re-generated at the onset of each nuclear cycle.

782
Movie 2 -A frame by frame 2D projection of movie 1 done using the ImSAnE tool (Heemskerk and Streichan, 2015). Dl-GFP 783 appears in grey.

784
Movie 3 -Time lapse of Dl-GFP intensity data for the area inside the dashed frame in Figure 1C. Each circular marker in the 785 movie shows raw, non smoothed Dl-GFP intensity in a single nucleus. Nuclei were segmented from the corresponding frames

795
. CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.    CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.  . CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. . CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.

814
Our full model describes how the Spz gradient is formed by shuttling, induces the nuclear 815 localization of Dl and the negative feedback between WntD and Dorsal which maintains the 816 gradients robustness (Rahimi et al., 2016). To this end, we extend the model from our previous 817 paper ( Rahimi et al., 2016) to include the nuclear localization of Dorsal. Also, we used a 818 different mechanism by which WntD contracts the Spz gradient: instead of competing with the 819 ligand for binding the Toll receptor, we assume here that WntD "crowds" the Toll receptors 820 immediate environment by binding its own receptor Frizzled4 and limiting the access of Spz to 821 Toll. This increases the chances of free ligand binding the shuttling molecule instead of the 822 receptor and therefore enhances shuttling. This "crowding" of Toll has an additional affect: 823 stabilizing the ligand which succeeded in binding, which also makes shuttling more efficient. 824 The governing set of reaction-diffusion equations of our model is given below: eqn. 1 defines 825 the temporal dynamics of freely diffusing WntD. The terms of the equation by order of 826 appearance describe: WntD diffusion, WntD degradation, WntD production which depends on 827 nuclear Dorsal. This last term is the induction part of InC as WntD production is positively 828 regulated by signaling. The WntD producing zone is restricted and is defined in the embryo by 829 the Torso signaling border. In the simulations we define this zone using the model parameter when bound together as the NC-Spz* complex. Signaling occurs when C-Spz bound to Toll 844 . CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint . http://dx.doi.org/10.1101/505925 doi: bioRxiv preprint first posted online Dec. 26, 2018; undergoes endocytosis (eqns. 5,8,10). Toll receptors undergo recycling back to the membrane 845 after endocytosis and the total concentration of Toll is constant (eqns. 10,3 respectively). The 846 C-Spz ligand and N-Spz inhibitor are products of NC-Spz complex separation when bound to 847 the Toll receptor (eqn. 4,9). NC-Spz is also capable of inducing Toll endocytosis when binding 848 it and thus contributing to signaling (eqn. 9) but signaling through NC-Spz happens at a much 849 lower rate than C-Spz mediated signaling. Equations 11-12 describe the induction of Dorsal 850 nuclear localization by Toll signaling. Eqn. 11 introduces the following constraint: the total 851 amount of Dorsal (Nuclear-and cytoplasmic-) is constant and equals . The 852 meaning of the different parameters and their units are summarized in Table S1. This set of 853 equations was solved numerically in 1D using a standard MATLAB PDE solver. 854  . CC-BY 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.