Kinetic Control of Out-Of-Equilibrium Dynamics in the RhoA Signaling Cascade Shapes Actomyosin Contractility

22 Cellular functions rely on the precise timing of signal transmission through sequential 23 activation cascades, yet the origin and functional role of signaling delays remain poorly 24 understood. Here, we dissect the temporal organization of the RhoA signaling cascade during 25 pulsed actomyosin contractility in the early C. elegans embryo. We uncover a stereotypical 26 delay between upstream RhoA/ROCK activation and downstream myosin II recruitment. Using 27 TIRF single-molecule microscopy, we show that this delay arises from binding and unbinding 28 kinetics of myosin rather than from slow biochemical reactions. A simple and versatile kinetic 29 model parameterized by these measurements accurately predicts the temporal evolution of 30 myosin accumulation and reveals active control of the dynamic range of the cascade . 31 Perturbing actin and myosin turnover experimentally confirms these predictions , and 32 numerical simulations show that the delay between actin and myosin plays a critical role in 33 force deployment during pulsed contraction. Together, our results indicate that kinetic delays 34 in signaling cascades are not simply tolerated during morphogenesis, but actively shape force 35 deployment in the cell. 36 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 3

40 Signaling cascades are a widespread biological strategy to respond to, propagate and 41 shape extracellular and intracellular inputs. Operating across multiple spatial and temporal 42 scales over a wide repertoire of molecular substrates –from Rho GTPases, to second 43 messengers, to transcriptional regulation– signaling cascades carry critical information within 44 biological systems (Novák and Tyson 2008; Kholodenko 2006; Ferrell 2013; Goldbeter 1996) . 45 The architecture and tuning of these cascades underlie a variety of behaviors, ranging from 46 signal magnitude or sensitivity amplification (Jr., Goldbeter, and Stock 1982) , ultrasensitivity 47 (Huang and Ferrell 1996) or bi-stability (Markevich, Hoek, and Kholodenko 2004; Mori, Jilkine, 48 and Edelstein-Keshet 2008), to cyclic oscillations (Pomerening, Sontag, and Ferrell 2003) or 49 chaotic dynamics (Gérard and Goldbeter 2012; Yamaguchi, Ode, and Ueda 2021) , and even 50 intricate dynamics within seemingly simple 3 -component networks (Ma et al. 2009) . While 51 mathematical models provide a glimpse on how these signaling cascades are tuned, and even 52 as optogenetic tools offer a handle to control cascade activation while monitoring its 53 propagation (Valon et al, 2015, de Seze et al, 2025), experimental measurements on dynamic 54 systems still remain notably constrained. The experimental dissection of the molecular 55 mechanisms underpinning the spatial and temporal dynamics of these activation cascades 56 therefore represent a key objective to understand how signals propagate in the cell. 57 To dissect the dynamics of hierarchical signaling cascades, we decided to focus on the 58 RhoA cascade, which orchestrates actomyosin recruitment during pulsed contractions. Small 59 RhoGTPases represent a key example of signaling cascades, yet the dynamics governing the 60 activation of the downstream players within these cascades remains largely uncharted. Over 61 the past 25 years, RhoA in particular has emerged as a central regulator of morphogenesis 62 (Hariharan et al. 1995; Eaton et al. 1995) , from Drosophila gastrulation, germband extension 63 and amnioserosa contraction during dorsal closure (Barrett, Leptin, and Settleman 1997; 64 Munjal et al. 2015; Solon et al. 2009), cells convergent extension in Xenopus early embryo (H. 65 Y. Kim and Davidson 2011), to compaction in the early mouse embryo (Maître et al. 2015). In 66 C. elegans, RhoA contributes to diverse processes including spermatheca contraction (Tan and 67 Zaidel-Bar 2015) , epidermal cell migration (Wallace et al. 2018) , embryonic elongation 68 (Diogon et al. 2007; Gally et al. 2009), and plays a critical role in embryo polarization (Motegi 69 and Sugimoto 2006; Schonegg et al. 2007) . At the molecular level, RhoA drives the dynamic 70 remodeling of cortical actomyosin networks and actomyosin contractility through a dual 71 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 4 effect on F-actin and Myosin II. RhoA activates and recruits the mDia/Dia formin homolog CYK-72 1, which processively elongates actin filaments (Kovar and Pollard 2004; Costache et al. 2022). 73 RhoA also activates and recruits the Rho kinase/ROCK (LET -502), which phosphorylates 74 Myosin Regulatory Light Chain and inhibits Myosin Phosphatase (MEL-11) (Diogon et al. 2007; 75 Piekny and Mains 2002; Wissmann et al. 1997; Wissmann, Ingles, and Mains 1999) resulting 76 ultimately in Myosin II activation and cortical recruitment (Fig. 1A). 77 Dynamically, RhoA local activation drives actomyosin activation leading to continuous 78 contraction of actomyosin cables, but also frequently occurs in pulses (Miao & Blankenship 79 2020). Pulsed actomyosin contractility represents a widespread mode of actomyosin 80 contractility, and drives a broad variety of morphogenetic processes, from cell and tissue 81 invagination during gastrulation to cell behavior coordination. Depending on the biological 82 context, the period of pulsed contractions varies widely, from 30 s in C. elegans (Michaux et 83 al. 2018), to 2 min in Drosophila gastrulation or germband extension (Martin, Kaschube, and 84 Wieschaus 2009; Munjal et al. 2015), to 6 min in Drosophila follicle cells (He et al. 2010). While 85 recent work has started to explore how the duration and iteration of contractions is linked to 86 deformation reversibility and effectiveness (Cavanaugh et al. 2020; Staddon et al. 2019) , the 87 molecular mechanisms underpinning signal transduction in this cascade still remain unclear, 88 in particular regarding the delay between steps of the cascade (Michaux et al. 2018) , the 89 molecular origins of these delays, and their impact on cellular mechanics. 90 Here, we first describe the time delay between consecutive steps of this hierarchical 91 cascade. Using single -molecule microscopy, we provide dynamic measurement in live 92 embryos of the binding rate (number of molecules binding to the cortex per time unit, noted 93 kapp) and the unbinding rate (fraction of molecules unbinding from the cortex per time unit, 94 noted koff), thereby exposing dynamic modulations of the signaling kinetics. Using a simple 95 mathematical model, we then explore numerically the impact of these modulations. Based on 96 this simple model, we derive the signaled concentration –a concentration that the system 97 would reach if the binding and unbinding rates were to remain constant– which monitors the 98 out-of-equilibrium dynamics of the system. We subsequently use our dynamic measurements 99 of kapp and koff to infer the signaled concentration in vivo . To modulate pulse period and 100 challenge our model, we then use genetic perturbations to modulate F -actin 101 assembly/disassembly and Myosin II binding/unbinding rates. Finally, we leverage computer 102 simulations to explore the impact of the delay on F -actin contraction. Together, our results 103 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 5 show that kinetic delay s are an integral part of signaling cascades and strongly impact 104 mechanical outputs, affecting the effective deployment of actomyosin contractility. 105 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 6

106 Pulsed contractions as model to explore signal propagation in a signaling cascade 107 The RhoA cascade display four key features that were of interest to us. First, it is highly 108 hierarchical, with two branches derived from the same upstream signal, a F-actin branch and 109 a Myosin II branch. Second, these two branches of the cascade contribute to generate the 110 output dynamics, raising the question of the importance and modalities of the synchronization 111 between two outcomes in parallel transduction pathways from a single input. Third, the 112 cascade displays a pulsatile activation, which we effectively used as iterative cycles of 113 activation to study signal propagation in the cascade. Finally, the RhoA signaling cascade is 114 interfaced with a mechanical output, raising generic questions regarding how mechanical and 115 biochemical signaling can intertwine. 116 In the zygote, however, pulsed contractions are associated with deep invaginations of 117 the cortex, potentially affecting measurements of pulse kinetics, hampering a detailed analysis 118 of pulse kinetics, in particular for single-molecule microscopy analysis. We therefore decided 119 to focus on the dynamics of the cascade at the 2 -cell stage. We have previously shown that 120 RhoA-driven pulsed activation causes local cortical actomyosin contractility (Michaux et al. 121 2018), potentially affecting the tension distributed in the actomyosin cortex and causing 122 mechanosensitive recruitment of cascade components. The amount of contraction itself, 123 however, remains reasonably small (~5%), arguably affecting only marginally the local 124 concentration of the components of the cascade. Finally, the RhoA signaling cascade unfolds 125 at the cell surface, greatly simplifying the observation. 126 127 Myosin II accumulation is delayed compared to RhoA activation 128 We first decided to proceed to an in-depth description of the temporal kinetics of the 129 RhoA signaling cascade. We deployed a general strategy to measure the dynamics of 130 accumulation of the sequential players of the activation cascade, focusing initially on the 131 Myosin II branch. Using strains co-expressing Myosin Heavy-Chain NMY-2 labeled with mKate2 132 (NMY-2::mKate2, (Dickinson et al. 2017) ) along with GFP -tagged probes for each individual 133 player of the cascade, we observed the cell cortex with near -total internal reflection (TIRF) 134 microscopy to assess Myosin II accumulation and pulse initiation (Fig. 1A). We visually 135 identified pulses as local accumulations of Myosin II, spanning several microns, in the anterior 136 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 7 (AB) cells during interphase at the 2 -cell stage, then quantified and compared the 137 accumulation of Myosin II with the accumulation of the GFP-tagged probe (Fig. 1B, C). 138 To establish a baseline control for our measurements, we used a strain co -expressing 139 on the one hand NMY -2 fused with mKate2 (red fluorescent protein) at the genomic locus, 140 and on the other hand an ectopic copy of NMY -2 fused with GFP inserted in the genome 141 (Nance 2003) . Simple observation revealed that both fusion proteins were expressed in 142 overlapping patterns, with almost identical spatial distributions around the center of the pulse 143 (Movie S1). We then measured the accumulation kinetics of NMY-2::GFP and NMY-2::mKate2. 144 As expected, we observed on average no significant overall difference in (Fig. 1D, Fig. S1), 145 confirming the validity of our approach. Interestingly, we observed that most delays were 146 comprised within a ~1 s window, thus establishing an upper range for measurement errors 147 (Fig. 1D, Fig. S1E, E’). 148 To compare the dynamics of accumulation of Myosin II with RhoA, we then used a 149 RhoA biosensor derived from C -terminus of Anillin fused to GFP to monitor RhoA 150 accumulation (hereon, GFP::AHPH, (Tse et al. 2012)). As previously described (Michaux et al. 151 2018), we observed a strong spatial and temporal correlation between GFP::AHPH and NMY -152 2::mKate2, with a median delay of ~4.5 s of NMY -2::mKate2 with respect to GFP::AHPH (Fig. 153 1D-E, Movie S2, Fig. S2). These results however only reported on the dynamics of a RhoA 154 biosensor, thus potentially inserting an additional intermediary step in the observed kinetics. 155 To observe directly a sequence of recruitment, we focused on the next step of the 156 cascade. Myosin II activation depends on the phosphorylation level of the Regulatory Light -157 Chain, which is controlled by phosphorylation by ROCK and dephosphorylation by the Myosin 158 Phosphatase (Karess et al. 1991) . We therefore compared the respective dynamics of ROCK 159 and Myosin II, and used a strain coexpressing NMY-2::mKate2 with Rho Kinase fused with GFP 160 at the endogenous genomic locus (GFP::LET-502, (K. R. Bell et al. 2020)). As for GFP::AHPH, we 161 observed a coupled, delayed accumulation of GFP::LET-502 with NMY-2::mKate2. Importantly, 162 we measured a delay of ~4.5 s between ROCK and Myosin II, not significantly different from 163 the delay between RhoA and Myosin II (Fig. 1D-E, Fig. S1A-D’, Movie S3). 164 As the inhibition of the Myosin Phosphatase by ROCK has been proposed to contribute 165 to the activation of Myosin II (Piekny and Mains 2002) , we also decided to monitor the 166 recruitment dynamics of the MEL -11 regulatory subunit of myosin phosphatase, and 167 generated a MEL -11::GFP CRISPR knockin. In C. elegans, mel-11 mutants display a range of 168 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 8 phenotypes, demonstrating that mel-11 plays multiple key roles at several stages during 169 embryonic morphogenesis (Piekny and Mains 2002; Wissmann, Ingles, and Mains 1999) . As 170 our CRISPR strain displayed no functional phenotype, we concluded that the fusion protein 171 was functional. In early embryos, MEL -11::GFP was cortically enriched, and displayed a 172 dynamic, grainy distribution at the cell surface, accumulating in particular at the cleavage 173 furrow during cell division (Movie S3), as previously proposed based on immunostaining 174 (Piekny and Mains 2002) . Co -expressing NMY -2::mKate2 with MEL -11::GFP, we compared 175 their accumulation dynamics. During pulsed contractions at the 2 -cell stage, MEL-11 showed 176 dynamic recruitment at the cortex, which did not visually seem to match pulsed contractions 177 (Movie S4). As expected from this observation, attempting to measure the delay between 178 NMY-2::mKate2 and MEL-11::GFP resulted in very poorly synchronized curves (Fig. S1G, Movie 179 S4) and broad dispersion of the delay measurements (Fig. S1G’), further supporting the 180 absence of coordination between MEL-11::GFP cortical recruitment and pulsed contractions. 181 Interestingly, this dispersion differed strongly from the dynamics we observed for all the other 182 players of the activation cascade (Fig. 1D, Fig. S1A-F’), suggesting that Myosin II accumulation 183 is controlled locally by its activation through RhoA and ROCK while the regulation of Myosin II 184 by its phosphatase is not a local process, but instead takes place at the scale of the embryo. 185 In previous experiments, we used Myosin Heavy -Chain fusion NMY -2::mKate2 as a 186 reporter of Myosin II dynamics. We however wondered if we could distinguish the 187 accumulation dynamics of the Heavy-Chain and Myosin Regulatory Light-Chain, and identify a 188 kinetic signature suggesting a dynamic exchange of the Myosin Regulatory Light -Chain from 189 the Myosin Heavy-Chain. To address this question, we generated a Myosin Regulatory Light -190 Chain MLC -4 GFP CRISPR knock -in strain. Previous work showed that mlc-4 is required for 191 proper embryogenesis (Shelton et al. 1999) . As the knock -in strain displayed no overt 192 phenotype, we concluded that MLC -4::GFP was functional. In early embryos, MLC -4::GFP 193 displayed a distribution very similar to that of NMY-2::GFP and NMY-2::mKate2. Co-expressing 194 MLC-4::GFP with NMY -2::mKate2, we observed that MLC -4::GFP, and NMY -2::mKate2 were 195 recruited with extremely similar spatial and temporal dynamics, actually reminiscent of our 196 NMY-2::GFP/NMY-2::mKate2 control experiment (Fig. 1D, Fig. S1E-F’, Fig. S2E-F). As 197 previously for NMY -2::GFP and NMY -2::mKate2, we could not measure a statistically 198 significant delay between the recruitment of the two chains (Fig. 1D, Fig. S1F -F’, Movie S4). 199 This observation supports the idea that Myosin II is recruited to the cortex as a hexamer, and 200 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 9 that the Myosin Regulatory Light -Chain, once assembled, does not display an exchange 201 dynamics on the Myosin Heavy-Chain, or that this exchange takes place at a timescale outside 202 of our temporal resolution. 203 We thus dissected the signaling cascade that leads to the recruitment of Myosin to the 204 cortex, precisely defining the delay between Rho, ROCK and Myosin. Importantly, our results 205 looking at Myosin, either with NMY -2::GFP and MLC -4::GFP, show that measurement errors 206 are rather small, and cannot explain the broader range of delays observed with either 207 RhoA/AHPH or ROCK. This variability, be it stochastic or not, shows that contractility shows 208 some level of robustness with respect to the observed range of delays. These observations 209 also raise the question of whether the timing relationships uncovered here are functionally 210 important for pulsed contractility, rather than being passive consequences of molecular 211 turnover. 212 213 Myosin II-ROCK accumulation delay results from Myosin II binding/unbinding kinetics 214 Taken together, our results demonstrated that the accumulation of Myosin II at the 215 cortex mirrored the accumulation of ROCK, with a time delay of 4.5 s. It remained unclear, 216 however, whether the delay originated from the kinetics of the phosphorylation reaction or if 217 it resulted from the recruitment dynamics of Myosin II. In order to clarify the molecular 218 mechanisms underlying the time delays that we observed in the cascade, we decided to focus 219 on the kinetics of Myosin II accumulation. 220 We thus performed single-molecule microscopy with particle tracking. We previously 221 established this method to explore kinetics and mobility of fusion proteins expressed at single-222 molecule levels in the early C. elegans embryo (Robin et al. 2014; Michaux et al. 2018) . To 223 visualize the dynamics of individual molecules during pulsed contractions, we used an 224 overexpression strain carrying NMY -2 fused with GFP over an endogenous NMY -2 225 background, and used RNAi against the GFP to specifically decrease the expression of the 226 NMY-2::GFP fusion protein, reverting the initial over -expression to a wild -type phenotype 227 with minute levels of NMY-2::GFP (Fig. 2A-C, Movie S5). Using automated particle tracking, we 228 then tracked individual molecules, and measured the appearance, density and disappearance 229 of molecules in pulsed contractions (Fig. 2F). As in previous studies (Watanabe 2002; Vallotton 230 et al. 2004; Ponti 2004; Robin et al. 2014; Michaux et al. 2018), we assumed that appearance, 231 fraction of disappearing molecules, and density, reported directly on the local cortical binding 232 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 10 rate (hereon kapp) (Fig. 2D), unbinding rate ( koff) (Fig. 2E), and local density, respectively. 233 Finally, we previously showed that, provided that photobleaching rate is low compared to 234 turnover, tracking indeed provides a reasonable estimate of the turnover rate, a condition 235 which we verified for Myosin II (Robin et al. 2014; Michaux et al. 2018) . We also accounted 236 for the impact of local cortical shrinkage/expansion, using a routine to follow adaptive regions 237 of interest (Fig. 2C). Briefly, we used the position of molecules to infer local strain and 238 corrected the region of interest so as to proceed to our quantifications in an adaptive frame 239 of reference, thereby removing confounding effects of advection of the cortex (for details, see 240 Michaux et al., 2018). These effects, however, only affect the measurements in a very limited 241 manner (<5%). 242 To explore our data, we further hypothesized a very simple kinetic accumulation 243 model, very similar to weakly activated cascades model ((Heinrich, Neel, and Rapoport 2002), 244 see Supplementary Information): 245 𝑑[𝑀𝑦𝑜∗] 𝑑𝑡 ≈ 𝑘𝑜𝑛 ∙ [𝑅𝑂𝐶𝐾] ∙ [𝑀𝑦𝑜𝑇] ∙ (1 − [𝑀𝑦𝑜∗] [𝑀𝑦𝑜𝑇]) − 𝑘𝑜𝑓𝑓 ∙ [𝑀𝑦𝑜∗] 246 where [ Myo*] is the concentration of phosphorylated Myosin II, [ MyoT] is the total 247 concentration of Myosin II. Under the assumption that the amount of activated Myosin II does 248 not deplete the cytoplasmic stock ([ Myo*] ≪ [MyoT]), and that ROCK recruitment is not 249 affected by Myosin II, we can write: 250 𝑑[𝑀𝑦𝑜∗] 𝑑𝑡 ≈ 𝑘𝑜𝑛 𝑎𝑝𝑝 − 𝑘𝑜𝑓𝑓 ∙ [𝑀𝑦𝑜∗] 251 where the effective binding rate (𝑘𝑜𝑛 𝑎𝑝𝑝, or simply kapp) and the unbinding rate (koff) determine 252 the evolution of the cortical concentration of Myosin II (Fig. 3B). Interestingly, under the 253 previous assumptions, kapp is linearly dependent on ROCK concentration. 254 Using this equation, and dynamic measurements of effective binding and unbinding 255 rates, we can define a n instantaneous steady-state concentration [Myo0] for each couple of 256 kapp, koff, such that [ Myo0] = kapp/koff. Biologically, this concentration represents the steady -257 state concentration that the system would eventually reach if kapp and koff were to remain 258 constant over time, which we call hereon cortical signaled concentration . Importantly, the 259 signaled concentration thus reflects the local signaling intensity of the immediate upstream 260 regulator in the signaling cascade, here the concentration of active ROCK. Fundamentally, 261 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 11 comparing the signaled and effective concentrations actually informs on how far out of 262 equilibrium the system lies. 263 This experimental signaled concentration value therefore integrates modulations of 264 binding and unbinding rate to reflect an activation intensity of the upstream signal in the 265 system. Comparing signaled concentration with effective concentration, we could therefore 266 measure a delay between the system and its target value, of approximately 4.25 s (Fig. 2H). 267 We then compared the delay between ROCK and Myosin II, measured in the previous section, 268 and the delay from the binding/unbinding kinetics, extracted from our single –molecule 269 measurements. Strikingly, the two results were extremely close, showing that the core of the 270 delay between the accumulation of ROCK and Myosin II essentially (4.25 s out of 4.5) results 271 directly from Myosin II binding/unbinding kinetics (Fig. 2G). 272 As the signaled concentration should reflect the effect of local ROCK activity, we 273 decided to compare signaled concentration with ROCK/LET-502 local relative intensity. To this 274 end, we aligned the average Myosin II effective concentration of our single-molecule Myosin II 275 dataset with the average relative intensity of NMY -2::mKate2 of our 2 -color GFP::LET -276 502/NMY-2::mKate2 dataset, thus synchronizing the two datasets (Fig. 2H). We then 277 compared the dynamics of the Myosin II signaled concentration with ROCK accumulation. We 278 observed that the two curves are highly overlapping, suggesting that our simple model 279 captures fairly accurately the dynamics of the system. 280 In essence, our results thus showed that the cortical recruitment of ROCK virtually 281 immediately modulates Myosin II kinetics. In contrast, the evolution of Myosin II 282 concentration towards a dynamically modulated Myosin II signaled concentration, 283 experimentally computed from kapp and koff, takes place with a time constant of several 284 seconds, in a manner resembling a capacitor’s charge. 285 286 Exploring the effect of binding/unbinding kinetics using numerical simulations 287 To better understand how modulations of upstream RhoA/ROCK dynamics affected 288 Myosin II cortical concentration, we turned to numerical simulations. In the cell, RhoA 289 accumulation is sometimes pulsed and unsynchronized, but also sometimes resembles a 290 pseudo-periodic sine wave (Fig. 3A). To get a sense of the effect of the modulation, by RhoA 291 and ROCK, of the binding and unbinding rates, we decided to simulate kapp and koff as sinusoid 292 modulations (Fig. 3B). Setting binding and unbinding rates, we could then compute both 293 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 12 effective and signaled concentrations, and numerically explore how these modulations 294 affected the evolution of the system. 295 Upon sinusoidal modulations of the binding rate and constant unbinding rate, we 296 observed the emergence of a delay between effective and signaled concentrations (Fig. 3C, 297 blue and red curves, resp.). Unsurprisingly, this delay decreased with increasing koff (Fig. S3A-298 C), the unbinding rate acting as a capacitor shifting the phase of the input signal, while the 299 output concentration remained sinusoidal. 300 An analytical solution of this problem (Supplementary Note 1), interestingly showcases 301 the high-pass filter effect of myosin turnover over upstream pulsed signals , with an 302 attenuation factor √1 + 𝜔2 𝑘𝑜𝑓𝑓2 . 303 Conversely, upon constant binding rate and sinusoidal modulations of the unbinding 304 rate, we observed that both effective and signaled concentration displayed a sharp periodic 305 peak around the time koff was minimal, and were essentially flat and overlapping as koff was 306 high (Fig. 3D). Interestingly the shape of the curve observed in vivo closely resembled this 307 second situation, with kapp constant, varying koff (Fig. 2G and 3D). Comparing the numerical 308 simulations of the model to the single -molecule results thus suggested that unbinding rate 309 modulation actually played a significant role in shaping Myosin II accumulation. 310 311 Perturbing Myosin II phosphorylation dynamics affects the dynamic range of the system 312 To test this model, we decided to perturb the unbinding dynamics of Myosin II using 313 mel-11(it26), a temperature -sensitive allele that behaves as a null allele at the restrictive 314 temperature that displays a hypercontractile phenotype during zygote polarization and early 315 cell divisions (Piekny and Mains, 2002). We crossed mel-11(it26) with a strain expressing NMY-316 2::GFP, imaged 2 -cell stage embryos at single -molecule levels as described in the previous 317 section, following the quantitative analysis as before (Fig. 2I–L). As expected from its 318 biochemical activity, mel-11(it26) displays an increased accumulation of cortical Myosin II 319 (Fig. 2L). We then observed that the average kapp increased ~3-fold compared to control, while 320 koff strikingly remained in the same range. Focusing on the dynamics of kapp and koff, we noted 321 that the relative amplitude of the variations in time of both variables, while still present, were 322 much milder in mel-11(it26) compared to control embryos (Fig. 2I–F, J –K). Strikingly, the 323 maximum-to-minimum ratio of the signaled concentration was reduced in the mutant 324 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 13 context: in the control situation, the signaled concentration increased ~6 - to 7 -fold from 325 minimum to maximum signaled concentration, while in contrast, in the mutant, the minimum 326 to maximum signaled concentration displayed only 2- to 3-fold variation (Fig. 2G, L, red curves, 327 Fig. 2M). Similarly, the maximum-to-minimum ratio of the effective concentration was much 328 lower in the mel-11(it26) mutant background (~1.25 -fold), compared to control (~2-fold) 329 (Fig. 2G, L, blue curves). Our results show that mel-11(it26), while displaying a stronger cortical 330 contractility (Piekny and Mains 2002) and higher cortical Myosin II density, also display weaker 331 variations in cortical Myosin II density. 332 Taken together, these results show that Myosin II dephosphorylation by the Myosin 333 Phosphatase MEL-11 promotes the emergence of strong, well defined pulses. The absence of 334 Myosin Phosphatase thus shifts Myosin II activation out of the dynamic range of the cascade 335 into a saturated regime where Myosin II remains constantly activated. Interestingly however, 336 perturbing Myosin Phosphatase activity did not significantly affect the measured delay 337 between signaled and effective concentration (Fig. 2L), suggesting a degree of a robustness of 338 this delay to perturbations of the phosphatase activity. 339 340 Actin accumulation reflects modulations of polymerization/depolymerization dynamics, and 341 precedes myosin accumulation 342 During pulsed contractions, in parallel with Myosin II, RhoA regulates F-actin dynamics 343 by directly regulating the formin CYK -1/mDia, an efficient actin filament elongation factor, 344 locally increasing Actin assembly rates (Swan et al. 1998; Costache et al. 2022; Naganathan et 345 al. 2018) . In order to explore how the same activation input was transduced across two 346 branches of a cascade, we thus decided to finely describe the temporal dynamics of the F-actin 347 branch of the RhoA cascade. 348 We first focused on the formin CYK -1. We therefore generated a strain co -expressing 349 CYK-1 fused with GFP at the endogenous genomic locus (CYK -1::GFP, (Costache et al. 2022)), 350 with NMY-2::mKate2, using NMY-2::mKate2 to synchronize our pulses. Using our strategy to 351 quantify cascade kinetics, we measured a delay of ~4.5 s (Fig. 4A -B, Movie S6). Interestingly, 352 this delay was not significantly different from the RhoA/Myosin II and ROCK/Myosin II delays 353 (Fig. 4A, Fig. S1A–B’, D, D’). 354 At this point, we suspected that F-actin and Myosin II were synchronously recruited at 355 the cortex. To test this hypothesis, we used a strain co -expressing an Actin reporter –the F-356 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 14 actin-binding Calponin Homology domain of Utrophin, fused with GFP (hereon UTR::GFP)– and 357 Myosin II. Surprisingly, we measured a statistically significant delay of ~0.5 s between F-actin 358 and Myosin II, F -actin appearing briefly before Myosin II (Fig. 4A -B, Fig. S1C, C’, Movie S7). 359 Additionally, as we were not directly observing F -actin, and instead using UTR::GFP to 360 indirectly monitor F-actin accumulation, our results further pointed to the accumulation of F-361 actin predating the accumulation of Myosin II, with a delay strictly greater than 0.5 s. 362 To understand the kinetics underlying F-actin accumulation, we wanted to access Actin 363 assembly and depolymerization rates. To this end, we used a strain expressing Actin fused 364 with GFP at single -molecule levels (hereon Actin::GFP), and measured the density of F -actin 365 during pulsed contractions, F -actin binding rate (assembly) and unbinding rate 366 (depolymerization), and F -actin signaled concentration (Fig. 4C–F, (Michaux et al. 2018) ). 367 Interestingly, we observed that the ratio between the maximum and minimum binding rates 368 during a pulse was much larger for F -actin (~5 -fold increase, from ~0.5 to ~2.5, Fig. 4C) 369 compared to Myosin II (~2-fold increase, Fig. 2D). In contrast, the comparative variations in 370 koff displayed the same order of magnitude (Fig. 4D, 2E). This observation suggested that F -371 actin and Myosin II cortical densities are modulated in very different ways, with a regulation 372 relying largely on assembly for F-actin, while it relies largely on disassembly for Myosin II. 373 In previous work (Costache et al. 2022) , we showed that during pulsed contractions, 374 two populations of formins are recruited at the cortex: recruited formins, functionally inactive 375 and immobile, and elongating formins, which actively elongate F-actin and move ballistically 376 in the cortex. We then decided to compare single -molecule actin dynamics with the 377 accumulation of the population of elongating formins. To synchronize these F -actin kinetics 378 data with our data on pulse dynamics, we aligned Actin::GFP cortical density with the average 379 curve of our UTR::GFP/NMY -2::mKate2 movies. Similarly, we generated single -molecule 380 microscopy movies of formin fused with GFP (hereon CYK-1::GFP), and used our kinetics data 381 from our CYK -1::GFP/NMY-2::mKate2 data to synchronize our single -molecule dataset with 382 our 2-color dataset. Using these synchronized datasets, we then compared the dynamics of 383 the F-actin signaled concentration with the dynamics of the population of elongating formin 384 (Fig. 4G). As for Myosin II, we observed that the two curves readily overlapped, suggesting 385 that we had captured the essential features of F-actin accumulation. 386 We finally wondered if F -actin dynamics could be affected by local capping dynamics 387 mediated by the C. elegans capping protein CAP-1, which had previously been implicated in 388 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 15 pulse dynamics (Naganathan et al. 2018) . Using a strain expressing a translational fusion 389 between CAP-1 and GFP at the endogenous genomic locus (hereon CAP-1::GFP), we observed 390 that CAP-1 was cortically enriched and displayed pulsed accumulations among other features. 391 In a strain co-expressing CAP-1::GFP with NMY-2::mKate2, we observed a delay of ~0.5 s (Fig. 392 4A, Fig. S1H-H’, Movie S8)between CAP-1 and Myosin II, similar to the delay we had measured 393 between Myosin II and F -actin. This observation suggests that F -actin is dynamically capped 394 at the cortex after F-actin barbed-end are released by formins. 395 396 Slowing down F-actin dynamics affects the dynamic range of the system 397 To test the effect of actin turnover on pulse dynamics, we decided to perturb and try 398 to predict the impact on the kinetics in the cascade. To slow down actin turnover, we used 399 RNAi against the actin severing protein gene, cofilin unc-60. In order to clarify the effects of 400 Cofilin, we first performed unc-60(RNAi) in embryos expressing very low levels of a 401 translational fusion of actin with GFP, thus monitoring actin dynamics directly (Robin et al. 402 2014; Michaux et al. 2018) . Using single -molecule measurements of Actin::GFP, we first 403 showed that Cofilin knock -down slowed F -actin turnover (Fig. S4E). We then measured the 404 average pulse period, and observed a shift from a distribution of pulse periods tightly -405 centered around ~31 s in control embryos to a much broader distribution centered around 406 42 s in unc-60 RNAi (Fig. S4F). Altogether, these results showed that unc-60(RNAi) embryos 407 displayed a generally slower F-actin turnover and pulse dynamics. 408 We then combined unc-60(RNAi) with NMY-2::GFP single-molecule microscopy, as in 409 our previous experiments, and measured Myosin II pulse density (Fig. 4J,K), binding rate (kapp, 410 Fig. 4H), unbinding rate (koff, Fig. 4I) and signaled concentration (kapp/koff, Fig. 4K). As expected 411 from the previous experiment (Fig. S4F), we observed broader pulses and an increased pulse 412 period in unc-60(RNAi) compared to control (Fig. S4C-D). Interestingly, the unbinding rate only 413 displayed very mild variations compared to the control. As for the Myosin Phosphatase 414 mutant mel-11(it26), we observed that the ratio between the maximum and minimum 415 signaled concentration was reduced (Fig. 2G, 4K), demonstrating a decreased response to the 416 upstream signal in unc-60(RNAi). Strikingly, neither unc-60(RNAi) nor mel-11(it26) displayed 417 strong changes in the delay we measured between the rises of signaled concentration and 418 effective concentration. 419 420 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 16 Cascade activation delay is required for efficient contractility during pulsed contractions 421 Our results showed that the two pathways downstream of RhoA activation lead to a 422 recruitment of the actin polymerizing enzyme, initiating an F-actin assembly ~4.5 s before 423 Myosin II accumulation, eventually resulting in F-Actin accumulation predating Myosin II 424 accumulation by >0.5 s. We therefore wanted to know if this these experimentally observed 425 difference of timing was functionally important for pulsed contractility. However, exploring 426 this question experimentally was difficult, as perturbations would affect multiple aspects of 427 pulsed contractions. In order to more clearly distinguish the effects of the dynamics of the 428 cascade on contractility, we therefore turned to agent -based simulation, to separate the 429 respective effects of timing, duration, and rate of filament elongation. 430 Using our established computational model of actomyosin networks (Fig. S4A, (Jung, 431 Murrell, and Kim 2015)), we probed the effect on cortical mechanics of the delay between the 432 initiation of formin -induced F -actin elongation in cortex and the recruitment of Myosin II. 433 Using a cortex -like F-actin meshwork (20 μm × 20 μm × 100 nm), we simulated RhoA -driven 434 pulsed contraction by locally modulating the kinetics of Myosin II and F-actin elongation rates, 435 based on previous experimental measurements (Costache et al. 2022) . Specifically, to 436 reproduce formin activation, we increased the elongation rate of a fraction of the barbed ends 437 in the RhoA-activated region to 1.2 µm/s, and keeping the barbed ends in this state for ~10 s, 438 thus generating rapidly elongating Actin filaments and reaching ~12 µm in length. In parallel, 439 to reproduce myosin activation, we locally turned on Myosin II activity in the RhoA -activated 440 region for 15 s. 441 To test the impact of delayed Myosin II recruitment on network architecture and the 442 deployment of Myosin II generated forces, we then tuned the duration of the delay between 443 formin activation and Myosin II activation as a variable (Jung, Murrell, and Kim 2015) . We 444 observed that simultaneous activation of formins and Myosin II lead to high contraction of 445 both F-actin and Myosin II, and to the formation of long-lasting F-actin aggregates. In contrast, 446 delay at 5 s and 10 s lead to less contraction of the network, avoiding network collapse. To 447 better quantify the mechanical impact of the delay, we measured the network contraction for 448 F-actin and Myosin II (Fig. 5B, C), as well as the sum of the forces and the average force 449 generated locally on the network by the pulsed contraction (Fig. 5C, D). Using these metrics, 450 we observed that Myosin II and F-actin contraction indeed decreased with increasing delay 451 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 17 (Fig. 5B, C). Interestingly, we also observed that the force deployed by the contraction evolved 452 non-monotonously and seemed higher when the delay was set at 5 s (Fig. 5D, E). 453 The case without delay showed larger maximum F -actin contraction presumably 454 because there was shorter time for elongated filaments to form cross -linking points before 455 Myosin II motors actively contract Actin filaments. Consistently, the sum of force acting on the 456 network was minimal in the case without the delay (Fig. 5D, E). From these observations, we 457 concluded that the relative timing of Myosin II accumulation and F-actin assembly is actually 458 linked to the contraction and force deployment in the network. 459

460 The RhoA pathway driving actomyosin contractility is an evolutionary conserved 461 signaling cascade that drives a broad range of morphogenetic processes, from cytokinesis to 462 cell-cell rearrangements and apical cell constriction to smooth muscle contraction. Depending 463 on the cellular context and RhoA activation modalities, how this signaling cascade is patterned 464 and unfolds can be highly variable, ranging from highly localized, pulsed dynamics to global 465 steady-state activation. Here, coupling live microscopy with image analysis, we identi fy the 466 enzymatic and kinetic bases that underlie, in vivo , the temporal dynamics of the RhoA 467 signaling cascade leading to actomyosin contraction. 468 Using the RhoA activation cascade in the early C. elegans embryo, we followed the 469 temporal dynamics of the cortical accumulation of successive components of cascade. 470 Interestingly, formin, ROCK and RhoA essentially display identical dynamics and delays, 471 suggesting RhoA effectors/binding partners accumulate with similar kinetics. This suggests 472 that our RhoA proxy –like Myosin II– is subject to a binding/unbinding delay and likely lags 473 behind RhoA, and f urther exploration of the kinetics of AHPH using single -molecule 474 microscopy and particle tracking –as we did for Myosin II and F-actin– should help resolve the 475 actual accumulation kinetics of active RhoA. 476 Surprisingly, while both RhoA effectors CYK-1 and ROCK/LET -502 accumulate with 477 similar dynamics, F-actin and Myosin II did not appear synchronously at the location of pulsed 478 contractions, F-actin preceding Myosin II by a small yet statistically significant ~0.5 s delay. 479 Physiologically, this can result from the fact that Myosin II cortical recruitment is dependent 480 on the presence of Actin filaments as binding substrate. In our quantitative analysis, this 481 aspect of Myosin ̛II dynamics was implicitly included in the measurement of kapp. While th is 482 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 18 delay is statistically significant on average , individual pulses often display an inversion of 483 recruitment timing . If indeed these inversions reflect genuine biological variability , they 484 suggest that the mechanical output of pulsed contractions is robust to modest variations in 485 the timing of Actin and Myosin recruitment. 486 Using our single -molecule data analysis, we established a simple model to extract a 487 dynamic signaled concentration value –in essence the instantaneous signaling activity of the 488 system– from time-resolved measurements of kapp and koff. Comparing this signaled 489 concentration with effective concentration lets us explore several interesting key properties 490 of the system, and in particular delays, signaling intensity, and out -of-equilibrium dynamics. 491 Importantly, this analysis showed that a single upstream RhoA pulse is decoded into two 492 distinct temporal signals: a rapid polymerization -based response in the actin network, and a 493 delayed, integrative contractile response mediated by myosin turnover. 494 Importantly, this kinetic model is not specific to the RhoA pathway, and could translate 495 to any number of signaling cascade , as for example MAP kinase cascades operating in the 496 weakly-activated regime (Heinrich, Neel, and Rapoport 2002; Beguerisse -Díaz, Desikan, and 497 Barahona 2016). 498 Specifically, the model provided a good theoretical framework to understand the 499 molecular bases for the delays we observed along the steps of the cascade. In particular, we 500 could show that Myosin II cortical accumulation relies essentially on the equilibration 501 dynamics under constantly changing kapp and koff. Numerical simulations of RhoA periodic 502 variations and its impact on kapp and koff, independently, showed that koff modulations under a 503 constant kapp resulted in dynamics extremely closer to the one we observed (Fig. S3). 504 In order to challenge this model, we experimentally perturbed Actin turnover rate, by 505 depleting the severing factor Cofilin with unc-60(RNAi). While, as expected, this caused an 506 increase in the pulsed contraction period, and abolished koff variation, it did not affect the 507 kinetic delay for Myosin II inferred from single -molecule measurements. This suggests that 508 Myosin II unbinding is effectively unaffected by Actin turnover. In this context, it therefore 509 seems that Myosin II unbinding operates independently from Actin filament severing. 510 To further challenge our model, we experimentally perturbed Myosin II 511 dephosphorylation by depleting the Myosin Phosphatase with mel-11(RNAi). Surprisingly, this 512 did not cause a decrease of the average Myosin II unbinding rate. We interpret this result by 513 the fact that Myosin II actually is a low/intermediate duty ratio motor that works 514 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 19 cooperatively as mini -filaments. Therefore, the Myosin II unbinding rate might simply 515 represent the unbinding rate of fully phosphorylated Myosin II. Under this assumption, we 516 would expect that an decreased phosphatase activity translates superficially in an increased 517 “available pool” of almost fully phosphorylated Myosin II that could be recruited at the cortex 518 by a weaker phosphorylation activity, thus exclusively affecting the magnitude of the binding 519 rate. This is consistent with the fact that let-502(sb106RNAi); mel-11(it26) double mutants are 520 viable: the hypomorph (non-fully functional) let-502 mutants would in this context work on a 521 larger pool of almost fully phosphorylated myosin, thus resulting in a balance between the 522 two phenotypes. Furthermore, this is consistent with a global, and slow, Myosin Phosphatase 523 activity, which matches our experimental observations regarding MEL -11::GFP spatial 524 localization. 525 Interestingly, the model also gave indications regarding the molecular bases 526 underlying the magnitude of the response in this dynamic system. In particular, we observed 527 that the RhoA cascade drove myosin concentration in regimes lying far out of equilibrium, 528 with a signaled concentration tuned very far from the effective concentration. In contrast, in 529 Myosin Phosphatase mel-11(RNAi) knockdown conditions, the dynamics of the system was 530 much more stable biochemically, suggesting that the signaling cascade likely operated outside 531 of the dynamic range of downstream myosin. The response time of the actomyosin system 532 thus acts as a filter in the time domain on the period of the upstream activation signal: slowly 533 evolving signals keep the system unperturbed, close to the steady state, while fast 534 modulations of the input signal cause the system to oscillate, wandering far out of equilibrium. 535 This behavior is reminiscent of a high -pass filter common in optics, electronics or mechanics, 536 and clearly visible in the analytical resolution of our simplified model representing the system 537 as a forced oscillator under sinusoidal driving signal. 538 Finally, our agent-based simulations shed some light on how the synchronization of F-539 actin and Myosin II could actually interplay and subsequently affect Myosin II contractility. We 540 could show that the delay between F -actin and Myosin II may play a physiologically relevant 541 role by promoting long range actomyosin contractility. In this specific context, agent -based 542 simulations of cortical mechanics thus offer an attractive alternative to active gel 543 hydrodynamic models, as they are well suited to capture this type of specifics of emergent 544 properties relying on the detailed architecture of the network. 545 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 20 This work thus reveals an unexpected aspect of signaling cascades, at the crossroads 546 of signaling dynamics, assembly of cytoskeletal architectures, and the mechanical properties 547 of actomyosin networks : kinetic parameters traditionally viewed as passive —binding and 548 unbinding rates—can themselves act as active signal-processing elements to determine both 549 the timing and mechanical outcome of cellular signaling. 550 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 21

551 552 Strains 553 List of strains used in this study can be found in Supplementary Table 4. Several strains 554 were provided by the CGC, which is funded by NIH Office of Research Infrastructure Programs 555 (P40 OD010440). MEL -11::GFP transgenic strain was generated by CRISPR as a translational 556 fusion with GFP(65C) at the C -terminus of MEL -11, using a short linker peptide 557 (ACCAGTGGTAGCGGC). CAP -1::GFP transgenic strain was generated by CRISPR as a 558 translational fusion with GFP(65C) at the C -terminus of CAP -1. This strain presents no overt 559 deleterious phenotype, suggesting that the fusion is functional. MLC-4::GFP CRISPR strain is 560 described in Suman, Chauca et al . Additional information, including primers and sg are 561 available from the authors upon request. 562 563 C. elegans culture, tools and RNAi 564 Unless specified otherwise, we cultured C. elegans at 20°C under standard conditions 565 (Brenner 1974). 566 Wormbase was used to obtain general information regarding genome, sequences and 567 available tools (Sternberg et al, 2024). 568 FBR221 hermaphrodites (mel-11(it26) unc-4(e120) sqt-1(sc13)/mnC1 dpy-10(e128) unc-569 52(e444)II; zuIs45 [nmy -2::NMY-2::GFP + unc -119(+)] V) were raised at the permissive 570 temperature (15°C). L3 -L4 homozygous for it26 were selected based on phenotype. 571 (mel-11(it26); zuIs45[nmy -2::gfp]) were placed on Nematode Growth Media (NGM) plates 572 containing 2 mM IPTG for 40 h at 15 °C with HT115 E. coli strain expressing GFP RNAi (L4440 573 GFP RNAi construct), then 1 h prior to the imaging were placed at the restrictive temperature 574 (25 °C). 575 The RNAi clone targeting unc-60 was derived from the Ahringer RNAi library (Kamath & 576 Ahringer 2003) , and its sequence verified and the plasmid retransformed into HT115(DE3) 577 bacteria. L4 larval stage animals carrying the Myosin II overexpression transgene zuIs45[nmy-578 2::gfp] were placed on Nematode Growth Media (NGM) plates containing 2 mM IPTG for 24 h 579 at 20 °C with HT115 E. coli strain expressing GFP RNAi 40 h prior imaging the embryos. 580 Animals were then moved to a plate with 1:1 mix of HT115 E. coli strains expressing GFP RNAi 581 and unc-60 RNAi, and imaged 20 h later. 582 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 22 583 Microscopy 584 Embryos were mounted as described previously (Robin et al., 2014) on 3 -well epoxy 585 glass slides (epoxy approx. 20 µm think) under #1.5 coverslips in 2.5 μl of standard Egg Salts 586 containing approx. 100 uniformly sized polystyrene beads (18.7 ± 0.03 μm diameter; no. 587 NT29N; Bangs Laboratories), to achieve uniform compression of the embryos for subsequent 588 TIRF microscopy. 589 We performed HILO microscopy (Tokunaga, Imamoto, and Sakata-Sogawa 2008) on an 590 inverted Nikon Ti-E microscope, equipped with a motorized TIRF illuminator, a CFI Apo 1.45 –591 NA/100× oil-immersion TIRF objective (Nikon) and a Ti -ND6-PFS-S Perfect Focus unit. Laser 592 illumination at 488 nm from a 50 -mW solid-state sapphire laser (Coherent) was delivered by 593 fiber optics to the TIRF illuminator. Images were magnified by 1.5× and collected on an Andor 594 iXon3 897 EMCCD or and Photometrics Prime 95B camera, yielding a pixel size of 107 nm and 595 73nm, respectively. Microscope and data acquisition parameters were controlled with the NIS 596 software version 4.50. 597 598 Pulsed contractions imaging and image analysis procedures 599 We imaged at 30% of 90 mW for 488 nm and 561 nm, with 50 ms exposure and no delay 600 between frames, corresponding to an effective 100 ms between two consecutive frames of 601 the same channel. Laser angle was set at 65°. Room temperature was set between 19 and 602 20,5°C. After acquisition, we used ImageJ to average five consecutive frames, in order to 603 achieve 2 frames per second. 604 Individual pulses were readily identified in the red channel by the local accumulation of 605 NMY-2. We used ImageJ software (NIH Image, Bethesda, MD) to extract sub -regions 606 containing a single pulsed contraction, as described previously (Michaux et al. 2018; Costache 607 et al. 2022) . Individual movies were subsequently loaded and analyzed in Matlab version 608 R2018a. The average image intensity was measured at every time frame for each pulse; the 609 intensity was then normalized: 610 𝐼𝑛𝑜𝑟𝑚𝑎𝑙𝑖𝑧𝑒𝑑 = 𝐼 − 𝐼𝑚𝑖𝑛 𝐼𝑚𝑎𝑥 − 𝐼𝑚𝑖𝑛 611 where 𝐼 = mean (Image intensity), 𝐼𝑚𝑎𝑥 = maximal intensity, and 𝐼𝑚𝑖𝑛 = minimal intensity. 612 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 23 Maximum intensity of the red channel is considered as 1 and the immediate precedent 613 minimum is as 0, minimum taken in the 50 preceding frames. Data for the Myosin II intensity 614 were aligned at 0.45, where the slope is maximal and provides the most accurate measure of 615 pulse alignment to fix the time at 0 s for all the red curves. This information is propagated to 616 data from the other channel (green) (which is also normalized to get an intensity between 0 617 and 1). The average for each distribution (red and green) is plotted with its confidence 618 interval (95%). 619 For measuring pulse size, we selected manually 5 pulses per embryo and used ImageJ to 620 measure the intensity profile along a line across each pulse for both green and red channels. 621 Comparing pulse sizes, we collected the full width at half maximum (fwhm) and measured the 622 difference between fwhm in green and red. 623 624 Statistical analysis 625 We performed one -sample Student tests (t -tests) to measure the significance of each 626 delay between paired red and green curves. We used paired -sample Student tests on delays 627 from each channel to measure the significance of the difference between the two delays. *** 628 means p<0.001, **: p<0.01, *: p<0.05, ns: non-significant. 629 630 Single-molecule imaging and Myosin II turnover analysis 631 We performed single-molecule imaging as described previously (Robin et al. 2014). We 632 used RNAi against GFP to reduce the number of imaged fluorescent molecules. We imaged 633 single molecules using 5 % of 90 mW of 488 nm laser, with 500 ms exposure, and no delay 634 between frames. Laser angle was set to 65 °. Room temperature was maintained between 19 635 and 20,5 °C. 636 We used Matlab implementation of the Crocker -Grier algorithm by the Kilfoil lab for single -637 particle tracking (Crocker and Grier 1996; Pelletier et al. 2009) . We used the following 638 parameters for all experiments at single -molecule level: particle size, 3 pixels, maximal 639 displacement of the particle between two consecutive frames, 4 pixels and memory to link 640 trajectories in non -consecutive frames, 3 frames. This last parameter allowed the tracking 641 software to look for the particle over multiple frames ensuring robust particle tracking. 642 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 24 Subsequent image analysis was performed in Matlab. Location and time frame of each pulse 643 was loaded in Matlab. At the “seed time”, we defined a Region Of Interest (ROI) corresponding 644 to a convex polygon bounded by molecules detected in the ellipse previously defined in 645 ImageJ (convex envelope of all the particles present in the ROI at the seed time frame). We 646 then displaced this ROI according to the motions of the molecules at the vertices –if present– 647 or by extracting a local velocity field from the motion of the surrounding molecules in a radius 648 of 30 px if the “edge molecule” had disappeared. The ROI thus faithfully described dynamics 649 of the cortex during a pulsed contraction excluding effects from advection caused by the local 650 contraction/expansion of the cortex. We then counted the number of molecules in each 651 frame, the number of molecules appearances (binding rate, kon) and the fraction of molecules 652 disappearing between two consecutive frames (unbinding rate, koff). 653 Data was normalized to the minimum/maximum number of detected NMY -2 particles, then 654 aligned to 0.45, where the slope is maximal and provides the most accurate measure for pulse 655 alignment. The time of alignment is defined as time = 0 s and is set from the curve of number 656 of molecules and propagated to the curves of binding and unbinding rate. To avoid effects of 657 the beginning and the end of the movie on the binding and unbinding rate, data of number of 658 molecules, for each pulse has been truncated of its first and final three values (replaced by 659 NaN in Matlab). Average for each measurement was plotted with its confidence interval 660 (95%). Equilibrium state value is obtained by division of measured binding and unbinding 661 rates. 662 We performed the turnover analysis as previously described (Robin et al. 2014) . Under the 663 assumption of a single population of particles, the number of molecules N(t) within a defined 664 region over time is governed by: 665 𝑑𝑁 𝑑𝑡 = 𝑘𝑎𝑝𝑝 − 𝑘𝑜𝑓𝑓𝑁 666 667 ⟺ 𝑁𝑡+1 − 𝑁𝑡 = (𝑘𝑎𝑝𝑝 − 𝑘𝑜𝑓𝑓𝑁)𝑑𝑡 668 669 where kapp represents the number of molecules binding (appearing) in a given region per unit 670 time (in molecule/s) and koff represents the fraction of molecules unbinding per unit time (in 671 s-1). 672 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 25 Supposing two different equilibrium states, with distinct 𝑘𝑎𝑝𝑝 and 𝑘𝑜𝑓𝑓 (state i, 𝑘𝑎𝑝𝑝 𝑖 and 𝑘𝑜𝑓𝑓 𝑖 , 673 with equilibrium 𝑁𝑒𝑞 𝑖 = 𝑘𝑎𝑝𝑝𝑖 𝑘𝑜𝑓𝑓 𝑖 ). Then the evolution of the number of molecules between state 674 1 and state 2, following a step-change from {𝑘𝑎𝑝𝑝 1 , 𝑘𝑜𝑓𝑓 1 } to {𝑘𝑎𝑝𝑝 2 , 𝑘𝑜𝑓𝑓 2 }, is governed by: 675 𝑁(𝑡) = 𝑘𝑎𝑝𝑝 2 𝑘𝑜𝑓𝑓 2 + (𝑘𝑎𝑝𝑝 1 𝑘𝑜𝑓𝑓 1 − 𝑘𝑎𝑝𝑝 2 𝑘𝑜𝑓𝑓 2 ) ∙ 𝑒−𝑘2𝑡 = 𝑁𝑒𝑞 2 + (𝑁𝑒𝑞 1 − 𝑁𝑒𝑞 2 )∙ 𝑒−𝑘2𝑡 676 Note that, in our equations, we use concentrations and absolute molecule numbers 677 equivalently, as we account for contractility in tracking our regions of interest. A more generic 678 solution with sinusoidal forcing is proposed in Supplementary Note 1. 679 To evaluate the difference between signaled concentration and measured density, we 680 computed the fold-difference as follows: 681 ∆ = 𝑚𝑎𝑥𝑡𝑎𝑟𝑔𝑒𝑡 − 𝑚𝑖𝑛𝑡𝑎𝑟𝑔𝑒𝑡 𝑚𝑎𝑥𝑚𝑒𝑎𝑠𝑢𝑟𝑒𝑑 − 𝑚𝑖𝑛𝑚𝑒𝑎𝑠𝑢𝑟𝑒𝑑 682 683 Comparison between total intensity and dynamic variation at single-molecule level 684 Normalized average value for ROCK and Myosin II variation in total intensity were 685 extracted from the two -color imaging experiment (GFP::LET -502 and NMY -2::mKate2). The 686 average for Myosin II total intensity was aligned at 0.45 of the maximum with the normalized 687 value of the measured number of molecules in Myosin II single -molecule experiment. The 688 alignment was propagated to the kon, koff and signaled concentration of the same experiment. 689 On Fig. 2H, we displayed only Myosin II kon, koff and signaled concentration, and ROCK and 690 Myosin II average normalized intensity. 691 The normalized average value for formin and Myosin II variation in total intensity were 692 extracted from the two -color imaging experiment (CYK -1::GFP and NMY -2::mKate2). The 693 normalized average value for Myosin II total intensity in this experiment was aligned at 0.45 694 of the maximum with the normalized average value for Myosin II total intensity of the two -695 color imaging experiment for F -actin and Myosin II (UTR::GFP and NMY -2::mKate2). The 696 normalized average value for F-actin total intensity was aligned at 0.45 of the maximum with 697 the normalized value of measured number of molecules in F -actin single -molecule 698 experiment. The alignment was propagated to the kon, koff and the signaled concentration of 699 the same experiment. The normalized average value for formin total intensity in the two-color 700 imaging experiment was aligned at 0.45 of the maximum with the normalized number of 701 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 26 superdiffusive formin in formin single -molecule experiment (data published in Fig. 2A -B 702 (Costache et al. 2022). On Fig. 4G, we only displayed the average for F-actin kon and koff, F-actin 703 signaled concentration, normalized superdiffusive formin population and normalized actin 704 intensity. 705 706 Overview of the computational model of actomyosin mechanics 707 We used a well -established agent -based model of actomyosin networks based on the 708 Langevin equation (Li et al. 2017; Jung, Murrell, and Kim 2015; T. Kim et al. 2009; Mak et al. 709 2016). In the model, actin filament (F -actin), motor, and actin crosslinker proteins (ACPs) are 710 coarse-grained using cylindrical segments. The motions of all the cylindrical segments are 711 governed by the Langevin equation for Brownian dynamics. Deterministic forces in the 712 Langevin equation include bending and extensional forces that maintain equilibrium angles 713 formed by segments and the equilibrium lengths of segments, respectively, as well as a 714 repulsive force acting between neighboring pairs of segments for considering volume -715 exclusion effects. 716 The formation of F-actin is initiated by a nucleation event, followed by polymerization at the 717 barbed end and depolymerization at the pointed end. ACPs bind to F-actin without preference 718 for cross-linking angles at a constant rate and also unbind from F -actin at a force-dependent 719 rate determined by Bell’s law (G. I. Bell 1978). Each arm of motors binds to F-actin at a constant 720 rate, and it then walks toward the barbed end of F -actin or unbinds from F -actin at force -721 dependent rates determined by the parallel cluster model (Erdmann, Albert, and Schwarz 722 2013; Erdmann and Schwarz 2012) . For all simulations in this study, we used a thin 723 computational domain (20 × 20 × 0.1 μm) with periodic boundary conditions only in x and y 724 directions. In z direction, the boundaries of the domain exert repulsive forces on elements 725 that moved beyond the boundaries. At the beginning of each simulation, a thin actin network 726 is formed via self-assembly of F-actin and ACP. 727 For implementing RhoA activation, the domain is divided into 16 subdomains (4×4 in x and y 728 directions). Every 30 s, one of the subdomains is randomly selected and then activated. In the 729 activated subdomain, a fraction of the barbed ends of F-actins are randomly chosen and then 730 undergo faster polymerization by a factor, ρf, for the duration of τf. With the reference values 731 of ρf = 10 and τf = 10 s, F-actins are elongated by ~10 µm on average. After the time delay of 732 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 27 dM, motors in the activated subdomain are allowed to self -assemble into thick filament 733 structures for the duration of τM. The reference values of dM and τM are 5 s and 15 s, 734 respectively. These active motors in the form of thick filaments can contract the part of the 735 network in the activated subdomain. Once they become inactive after τM, the motors are 736 disassembled into monomers that cannot bind to F-actin. 737 738 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 28

739 Barrett, Kathy, Maria Leptin, and Jeffrey Settleman. 1997. “The Rho GTPase and a Putative 740 RhoGEF Mediate a Signaling Pathway for the Cell Shape Changes in Drosophila 741 Gastrulation.” Cell 91 (7): 905–15. doi:10.1016/s0092-8674(00)80482-1. 742 Beguerisse-Díaz, Mariano, Radhika Desikan, and Mauricio Barahona. 2016. “Linear Models of 743 Activation Cascades: Analytical Solutions and Coarse -Graining of Delayed Signal 744 Transduction.” Journal of The Royal Society Interface 13 (121): 20160409. 745 doi:10.1098/rsif.2016.0409. 746 Bell, George I. 1978. “Models for the Specific Adhesion of Cells to Cells.” Science 200 (4342): 747 618–27. doi:10.1126/science.347575. 748 Bell, Kathryn Rehain, Michael E. Werner, Anusha Doshi, Daniel B. Cortes, Adam Sattler, Thanh 749 Vuong-Brender, Michel Labouesse, and Amy Shaub Maddox. 2020. “Novel Cytokinetic Ring 750 Components Drive Negative Feedback in Cortical Contractility.” Molecular Biology of the 751 Cell 31 (15): 1623–36. doi:10.1091/mbc.e20-05-0304. 752 Brenner, S. 1974. “THE GENETICS OF CAENORHABDITIS ELEGANS.” Genetics 77 (1): 71 –94. 753 doi:10.1093/genetics/77.1.71. 754 Cavanaugh, Kate E., Michael F. Staddon, Edwin Munro, Shiladitya Banerjee, and Margaret L. 755 Gardel. 2020. “RhoA Mediates Epithelial Cell Shape Changes via Mechanosensitive 756 Endocytosis.” Developmental Cell 52 (2): 152-166.e5. doi:10.1016/j.devcel.2019.12.002. 757 Costache, Vlad, Serena Prigent Garcia, Camille N Plancke, Jing Li, Simon Begnaud, Shashi 758 Kumar Suman, Anne -Cécile Reymann, Taeyoon Kim, and Francois B Robin. 2022. “Rapid 759 Assembly of a Polar Network Architecture Drives Efficient Actomyosin Contractility.” Cell 760 Reports 39 (9): 110868. doi:10.1016/j.celrep.2022.110868. 761 Crocker, J C, and D G Grier. 1996. “Methods of Digital Video Microscopy for Colloidal Studies.” 762 Journal of Colloid and Interface Science 179 (1): 298–310. doi:10.1006/jcis.1996.0217. 763 de Seze J, Bongaerts M, Boulevard B, Coppey M. 2025. "Optogenetic control of a GEF of RhoA 764 uncovers a signaling switch from retraction to protrusion." Elife. 12:RP93180. doi: 765 10.7554/eLife.93180. 766 Dickinson, Daniel J., Francoise Schwager, Lionel Pintard, Monica Gotta, and Bob Goldstein. 767 2017. “A Single -Cell Biochemistry Approach Reveals PAR Complex Dynamics during Cell 768 Polarization.” Developmental Cell 42 (4): 416-434.e11. doi:10.1016/j.devcel.2017.07.024. 769 Diogon, M, F Wissler, S Quintin, Y Nagamatsu, S Sookhareea, F Landmann, H Hutter, N Vitale, 770 and M Labouesse. 2007. “The RhoGAP RGA -2 and LET -502/ROCK Achieve a Balance of 771 Actomyosin-Dependent Forces in C. Elegans Epidermis to Control Morphogenesis.” 772 Development 134 (13): 2469–79. doi:10.1242/dev.005074. 773 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 29 Eaton, S, P Auvinen, L Luo, Y N Jan, and K Simons. 1995. “CDC42 and Rac1 Control Different 774 Actin-Dependent Processes in the Drosophila Wing Disc Epithelium.” The Journal of Cell 775 Biology 131 (1): 151–64. doi:10.1083/jcb.131.1.151. 776 Erdmann, Thorsten, Philipp J. Albert, and Ulrich S. Schwarz. 2013. “Stochastic Dynamics of 777 Small Ensembles of Non -Processive Molecular Motors: The Parallel Cluster Model.” The 778 Journal of Chemical Physics 139 (17): 175104. doi:10.1063/1.4827497. 779 Erdmann, Thorsten, and Ulrich S. Schwarz. 2012. “Stochastic Force Generation by Small 780 Ensembles of Myosin II Motors.” Physical Review Letters 108 (18): 188101. 781 doi:10.1103/physrevlett.108.188101. 782 Ferrell, James E. 2013. “Feedback Loops and Reciprocal Regulation: Recurring Motifs in the 783 Systems Biology of the Cell Cycle.” Current Opinion in Cell Biology 25 (6): 676 –86. 784 doi:10.1016/j.ceb.2013.07.007. 785 Gally, Christelle, Fr édéric Wissler, Hala Zahreddine, Sophie Quintin, Fr édéric Landmann, and 786 Michel Labouesse. 2009. “Myosin II Regulation during C. Elegans Embryonic 787 Elongation:LET-502/ROCK, MRCK -1 and PAK -1, Three Kinases with Different Roles.” 788 Development 136 (18): 3109–19. doi:10.1242/dev.039412. 789 Gérard, Claude, and Albert Goldbeter. 2012. “Entrainment of the Mammalian Cell Cycle by the 790 Circadian Clock: Modeling Two Coupled Cellular Rhythms.” PLoS Computational Biology 8 791 (5): e1002516. doi:10.1371/journal.pcbi.1002516. 792 Goldbeter, Albert. 1996. Biochemical Oscillations and Cellular Rhythms: The Molecular Bases 793 of Periodic and Chaotic Behavior. Cambridge Univ. Press. Cambridge Univ. Press. 794 Hariharan, I K, K Q Hu, H Asha, A Quintanilla, R M Ezzell, and J Settleman. 1995. 795 “Characterization of Rho GTPase Family Homologues in Drosophila Melanogaster: 796 Overexpressing Rho1 in Retinal Cells Causes a Late Developmental Defect.” The EMBO 797 Journal 14 (2): 292–302. doi:10.1002/j.1460-2075.1995.tb07003.x. 798 He, Li, Xiaobo Wang, Ho Lam Tang, and Denise J. Montell. 2010. “Tissue Elongation Requires 799 Oscillating Contractions of a Basal Actomyosin Network.” Nature Cell Biology 12 (12): 800 1133–42. doi:10.1038/ncb2124. 801 Heinrich, Reinhart, Benjamin G. Neel, and Tom A. Rapoport. 2002. “Mathematical Models of 802 Protein Kinase Signal Transduction.” Molecular Cell 9 (5): 957 –70. doi:10.1016/s1097 -803 2765(02)00528-2. 804 Huang, C Y, and J E Ferrell. 1996. “Ultrasensitivity in the Mitogen -Activated Protein Kinase 805 Cascade.” Proceedings of the National Academy of Sciences 93 (19): 10078 –83. 806 doi:10.1073/pnas.93.19.10078. 807 Jr., Daniel E. Koshland, Albert Goldbeter, and Jeffry B. Stock. 1982. “Amplification and 808 Adaptation in Regulatory and Sensory Systems.” Science 217 (4556): 220 –25. 809 doi:10.1126/science.7089556. 810 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 30 Jung, Wonyeong, Michael P. Murrell, and Taeyoon Kim. 2015. “F-Actin Cross-Linking Enhances 811 the Stability of Force Generation in Disordered Actomyosin Networks.” Computational 812 Particle Mechanics 2 (4): 317–27. doi:10.1007/s40571-015-0052-9. 813 Kholodenko, Boris N. 2006. “Cell -Signalling Dynamics in Time and Space.” Nature Reviews 814 Molecular Cell Biology 7 (3): 165–76. doi:10.1038/nrm1838. 815 Kim, H Y, and L A Davidson. 2011. “Punctuated Actin Contractions during Convergent 816 Extension and Their Permissive Regulation by the Non-Canonical Wnt-Signaling Pathway.” 817 Journal of Cell Science 124 (4): 635–46. doi:10.1242/jcs.067579. 818 Kim, Taeyoon, Wonmuk Hwang, Hyungsuk Lee, and Roger D. Kamm. 2009. “Computational 819 Analysis of Viscoelastic Properties of Crosslinked Actin Networks.” PLoS Computational 820 Biology 5 (7): e1000439. doi:10.1371/journal.pcbi.1000439. 821 Kovar, David R, and Thomas D Pollard. 2004. “Insertional Assembly of Actin Filament Barbed 822 Ends in Association with Formins Produces Piconewton Forces.” Proceedings of the 823 National Academy of Sciences of the United States of America 101 (41): 14725 –30. 824 doi:10.1073/pnas.0405902101. 825 Li, Jing, Thomas Biel, Pranith Lomada, Qilin Yu, and Taeyoon Kim. 2017. “Buckling-Induced F-826 Actin Fragmentation Modulates the Contraction of Active Cytoskeletal Networks.” Soft 827 Matter 13 (17): 3213–20. doi:10.1039/c6sm02703b. 828 Ma, Wenzhe, Ala Trusina, Hana El -Samad, Wendell A. Lim, and Chao Tang. 2009. “Defining 829 Network Topologies That Can Achieve Biochemical Adaptation.” Cell 138 (4): 760 –73. 830 doi:10.1016/j.cell.2009.06.013. 831 Maître, Jean-Léon, Ritsuya Niwayama, Hervé Turlier, François Nédélec, and Takashi Hiiragi. 832 2015. “Pulsatile Cell-Autonomous Contractility Drives Compaction in the Mouse Embryo.” 833 Nature Cell Biology 17 (7): 849–55. 834 Mak, Michael, Muhammad H. Zaman, Roger D. Kamm, and Taeyoon Kim. 2016. “Interplay of 835 Active Processes Modulates Tension and Drives Phase Transition in Self-Renewing, Motor-836 Driven Cytoskeletal Networks.” Nature Communications 7 (1): 10323. 837 doi:10.1038/ncomms10323. 838 Markevich, Nick I., Jan B. Hoek, and Boris N. Kholodenko. 2004. “Signaling Switches and 839 Bistability Arising from Multisite Phosphorylation in Protein Kinase Cascades.” The Journal 840 of Cell Biology 164 (3): 353–59. doi:10.1083/jcb.200308060. 841 Martin, Adam C, Matthias Kaschube, and Eric F Wieschaus. 2009. “Pulsed Contractions of an 842 Actin-Myosin Network Drive Apical Constriction.” Nature 457 (7228): 495 –99. 843 doi:10.1038/nature07522. 844 Michaux, Jonathan B., François B. Robin, William M. McFadden, and Edwin M. Munro. 2018. 845 “Excitable RhoA Dynamics Drive Pulsed Contractions in the Early C. Elegans Embryo.” 846 Journal of Cell Biology 217 (12): 4230–52. doi:10.1083/jcb.201806161. 847 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 31 Mori, Yoichiro, Alexandra Jilkine, and Leah Edelstein -Keshet. 2008. “Wave -Pinning and Cell 848 Polarity from a Bistable Reaction -Diffusion System.” Biophysical Journal 94 (9): 3684–97. 849 doi:10.1529/biophysj.107.120824. 850 Motegi, Fumio, and Asako Sugimoto. 2006. “Sequential Functioning of the ECT -2 RhoGEF, 851 RHO-1 and CDC -42 Establishes Cell Polarity in Caenorhabditis Elegans Embryos.” Nature 852 Cell Biology 8 (9): 978–85. doi:10.1038/ncb1459. 853 Munjal, Akankshi, Jean -Marc Philippe, Edwin Munro, and Thomas Lecuit. 2015. “A Self -854 Organized Biomechanical Network Drives Shape Changes during Tissue Morphogenesis.” 855 Nature 524 (7565): 351–55. doi:10.1038/nature14603. 856 Naganathan, Sundar Ram, Sebastian Fürthauer, Josana Rodriguez, Bruno Thomas Fievet, Frank 857 Jülicher, Julie Ahringer, Carlo Vittorio Cannistraci, and Stephan W Grill. 2018. 858 “Morphogenetic Degeneracies in the Actomyosin Cortex.” ELife 7 (October): e37677. 859 doi:10.7554/elife.37677. 860 Nance, J. 2003. “C. Elegans PAR -3 and PAR -6 Are Required for Apicobasal Asymmetries 861 Associated with Cell Adhesion and Gastrulation.” Development 130 (22): 5339 –50. 862 doi:10.1242/dev.00735. 863 Novák, Béla, and John J Tyson. 2008. “Design Principles of Biochemical Oscillators.” Nature 864 Reviews Molecular Cell Biology 9 (12): 981–91. doi:10.1038/nrm2530. 865 Pelletier, Vincent, Naama Gal, Paul Fournier, and Maria Kilfoil. 2009. “Microrheology of 866 Microtubule Solutions and Actin -Microtubule Composite Networks.” Physical Review 867 Letters 102 (18): 188303. doi:10.1103/physrevlett.102.188303. 868 Piekny, Alisa J, and Paul E Mains. 2002. “Rho -Binding Kinase (LET -502) and Myosin 869 Phosphatase (MEL-11) Regulate Cytokinesis in the Early Caenorhabditis Elegans Embryo.” 870 Journal of Cell Science 115 (Pt 11): 2271–82. http://pubmed.gov/12006612. 871 Pomerening, Joseph R., Eduardo D. Sontag, and James E. Ferrell. 2003. “Building a Cell Cycle 872 Oscillator: Hysteresis and Bistability in the Activation of Cdc2.” Nature Cell Biology 5 (4): 873 346–51. doi:10.1038/ncb954. 874 Ponti, A. 2004. “Two Distinct Actin Networks Drive the Protrusion of Migrating Cells.” Science 875 305 (5691): 1782–86. doi:10.1126/science.1100533. 876 Robin, François B, William M McFadden, Baixue Yao, and Edwin M Munro. 2014. “Single -877 Molecule Analysis of Cell Surface Dynamics in Caenorhabditis Elegans Embryos.” Nature 878

11 (6): 677–82. doi:10.1038/nmeth.2928. 879 Schonegg, Stephanie, Alexandru T Constantinescu, Carsten Hoege, and Anthony A Hyman. 880 2007. “The Rho GTPase-Activating Proteins RGA-3 and RGA-4 Are Required to Set the Initial 881 Size of PAR Domains in Caenorhabditis Elegans One -Cell Embryos.” Proceedings of the 882 National Academy of Sciences of the United States of America 104 (38): 14976 –81. 883 doi:10.1073/pnas.0706941104. 884 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 32 Solon, Jerome, Aynur Kaya -Copur, Julien Colombelli, and Damian Brunner. 2009. “Pulsed 885 Forces Timed by a Ratchet-like Mechanism Drive Directed Tissue Movement during Dorsal 886 Closure.” CELL 137 (7): 1331–42. doi:10.1016/j.cell.2009.03.050. 887 Staddon, Michael F., Kate E. Cavanaugh, Edwin M. Munro, Margaret L. Gardel, and Shiladitya 888 Banerjee. 2019. “Mechanosensitive Junction Remodeling Promotes Robust Epithelial 889 Morphogenesis.” Biophysical Journal 117 (9): 1739–50. doi:10.1016/j.bpj.2019.09.027. 890 Sternberg PW, Van Auken K, Wang Q, Wright A, Yook K, Zarowiecki M, Arnaboldi V, Becerra A, 891 Brown S, Cain S, Chan J, Chen WJ, Cho J, Davis P, Diamantakis S, Dyer S, Grigoriadis D, Grove 892 CA, Harris T, Howe K, Kishore R, Lee R, Longden I, Luypaert M, Müller HM, Nuin P, Quinton-893 Tulloch M, Raciti D, Schedl T, Schindelman G, Stein L. 2024. “WormBase024: status and 894 transitioning to Alliance infrastructure.” Genetics 227 (1). doi:0.1093/genetics/iyae050. 895 Swan, K A, A F Severson, J C Carter, P R Martin, H Schnabel, R Schnabel, and B Bowerman. 896 1998. “Cyk-1: A C. Elegans FH Gene Required for a Late Step in Embryonic Cytokinesis.” 897 Journal of Cell Science 111 ( Pt 14) (July): 2017–27. http://pubmed.gov/9645949. 898 Tan, Pei Yi, and Ronen Zaidel -Bar. 2015. “Transient Membrane Localization of SPV -1 Drives 899 Cyclical Actomyosin Contractions in the C. Elegans Spermatheca.” Current Biology 25 (2): 900 141–51. doi:10.1016/j.cub.2014.11.033. 901 Tokunaga, Makio, Naoko Imamoto, and Kumiko Sakata -Sogawa. 2008. “Highly Inclined Thin 902 Illumination Enables Clear Single -Molecule Imaging in Cells.” Nature Methods 5 (2): 159–903 61. doi:10.1038/nmeth1171. 904 Tse, Yu Chung, Michael Werner, Katrina M Longhini, Jean -Claude Labbé, Bob Goldstein, and 905 Michael Glotzer. 2012. “RhoA Activation during Polarization and Cytokinesis of the Early 906 Caenorhabditis Elegans Embryo Is Differentially Dependent on NOP -1 and CYK -4.” 907 Molecular Biology of the Cell 23 (20): 4020–31. doi:10.1091/mbc.e12-04-0268. 908 Valon L, Etoc F, Remorino A, di Pietro F, Morin X, Dahan M, Coppey M. 2015. "Predictive 909 Spatiotemporal Manipulation of Signaling Perturbations Using Optogenetics." Biophys J. 910 109(9):1785-97. doi: 10.1016/j.bpj.2015.08.042. 911 Vallotton, Pascal, Stephanie L Gupton, Clare M Waterman -Storer, and Gaudenz Danuser. 912 2004. “Simultaneous Mapping of Filamentous Actin Flow and Turnover in Migrating Cells 913 by Quantitative Fluorescent Speckle Microscopy.” Proceedings of the National Academy of 914 Sciences of the United States of America 101 (26): 9660–65. doi:10.1073/pnas.0300552101. 915 Wallace, Andre G., Hamidah Raduwan, John Carlet, and Martha C. Soto. 2018. “The RhoGAP 916 HUM-7/Myo9 Integrates Signals to Modulate RHO -1/RhoA during Embryonic 917 Morphogenesis in Caenorhabditis Elegans.” Development 145 (23): dev168724. 918 doi:10.1242/dev.168724. 919 Watanabe, N. 2002. “Single -Molecule Speckle Analysis of Actin Filament Turnover in 920 Lamellipodia.” Science 295 (5557): 1083–86. doi:10.1126/science.1067470. 921 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 33 Wissmann, A, J Ingles, and P E Mains. 1999. “The Caenorhabditis Elegans Mel -11 Myosin 922 Phosphatase Regulatory Subunit Affects Tissue Contraction in the Somatic Gonad and the 923 Embryonic Epidermis and Genetically Interacts with the Rac Signaling Pathway.” 924 Developmental Biology 209 (1): 111–27. doi:10.1006/dbio.1999.9242. 925 Wissmann, A, J Ingles, J D McGhee, and P E Mains. 1997. “Caenorhabditis Elegans LET -502 Is 926 Related to Rho -Binding Kinases and Human Myotonic Dystrophy Kinase and Interacts 927 Genetically with a Homolog of the Regulatory Subunit of Smooth Muscle Myosin 928 Phosphatase to Affect Cell Shape.” Genes & Development 11 (4): 409 –22. 929 doi:10.1101/gad.11.4.409. 930 Yamaguchi, Hiroto Q., Koji L. Ode, and Hiroki R. Ueda. 2021. “A Design Principle for 931 Posttranslational Chaotic Oscillators.” IScience 24 (1): 101946. 932 doi:10.1016/j.isci.2020.101946. 933 934 935 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint RhoA Formin ROCK Myosin heavy-chainActin (proxy: GFP::AHPH) (CYK-1::GFP) (GFP::LET-502) (proxy: UTR::GFP) (NMY-2::mKate2) Myosin regulatory light-chain (MLC-4::GFP) A C D RhoA Formin ROCK MyosinActin ns ~4.5s >4.5s E B 0 105 15 20 time (s) 25 3530 40 AHPH/NMY-2 LET-502/NMY-2 NMY-2/NMY-2MLC-4/NMY-2 ns *** *** ns time (s) 0 -5 5 10 15 20 -10 GFP::LET-502 NMY-2::mKate2 GFP::LET-502 NMY-2::mKate2 Merge Figure 1 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint Figure 1. RhoA hierarchical signaling cascade reveals a delay between RhoA, ROCK and Myosin. (A) RhoA activation cascade of Actomyosin pulsed contractions with the different strains and proxys used in (D). (B) Near-TIRF microscopy image of a 2 -cell stage C. elegans embryo showing ROCK in green ( GFP::LET-502) and Myosin in magenta (NMY -2::mKate2), scale bar: 5 μm. (C) Timelapse of the pulsed contraction in white box in (A). (D) Quantifications of time delay. Blue dots: time delay distribution red over green channel, red dotted line: time delay of 0 s, blue shade: histogram of distribution. N(embryos) ≥ 10, N(pulses) ≥ 165 (See Supplementary Table S1 and S2 for statistical details), ns: not significant, ***: p-value ≤ 0.001. (E) Summary of the observed time delay within the RhoA activation cascade. .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 0.0 0.5 1.0 1.5 2.0 0 10 -10 20 time (s) ďŝŶĚŝŶŐ ƌĂƚe (mŽůeĐƵůe͘s-1) 0.0 0.2 0.4 0.6 0.8 1.0 0 10 -10 20 time (s) ŶŽƌmĂůŝnjeĚ ƉƵůse ĐŽŶĐeŶƚƌĂtiŽŶ 0.00 0.05 0.10 0.15 0 10 -10 20 time (s) ƵŶďŝŶĚŝŶŐ ƌĂƚe (s-1) 0 10 20 30 40 50 60 70 0 10 -10 20 time (s) ƉƵůse ĐŽŶĐeŶƚƌĂtiŽŶ A B C time (s) 0 10 5 35 15 20 25 30 ED F G H 0.0 0.2 0.4 0.6 0.8 1.0 0 10 -10 20 time (s) 30 -20ŶŽƌmĂůŝnjeĚ ǀĂůƵes 0 1 2 3 4 0 10 -10 20 time (s) ďŝŶĚŝŶŐ ƌĂƚe (mŽůeĐƵůe͘s-1) 0.00 0.05 0.10 0.15 0 10 -10 20 time (s) ƵŶďŝŶĚŝŶŐ ƌĂƚe (s-1) 0.0 0.2 0.4 0.6 0.8 1.0 0 10 -10 20 time (s) ŶŽƌmĂůŝnjeĚ ƉƵůse ĐŽŶĐeŶƚƌĂtiŽŶ 0 20 40 60 80 0 10 -10 20 time (s) ƉƵůse ĐŽŶĐeŶƚƌĂtiŽŶ I J K L 0 5 10 15 20 ƌĂtiŽ ŽĨ ĂmƉůŝƚƵĚes (sŝŐŶĂůeĚ Žǀeƌ eīeĐtiǀe) ŽŶƚƌŽůmel-11(it26) M *** ^ŝŶŐůeͲmŽůeĐƵůe mLJŽsŝŶ^ŝŶŐůeͲmŽůeĐƵůe mLJŽsŝŶ mel-11(it26) kapp DLJŽsŝŶ koff DLJŽsŝŶ DLJŽsŝŶ sŝŐŶĂůeĚ ĐŽŶĐeŶƚƌĂtiŽŶ ZK< /ŶƚeŶsŝƚLJ DLJŽsŝŶ eīeĐtiǀe ĐŽŶĐeŶƚƌĂtiŽŶ meĚŝĂŶ ĂǀeƌĂŐe meĚŝĂŶ ĂǀeƌĂŐe meĚŝĂŶ ĂǀeƌĂŐe meĚŝĂŶ ĂǀeƌĂŐe meĚŝĂŶ ĂǀeƌĂŐe meĚŝĂŶ ĂǀeƌĂŐe meĚŝĂŶ ĂǀeƌĂŐe īeĐtiǀe ĐŽŶĐeŶƚƌĂtiŽŶ ^ŝŐŶĂůeĚ ĐŽŶĐeŶƚƌĂtiŽŶ meĚŝĂŶ ĂǀeƌĂŐe meĚŝĂŶ ĂǀeƌĂŐe meĚŝĂŶ ĂǀeƌĂŐe Figure 2 īeĐtiǀe ĐŽŶĐeŶƚƌĂtiŽŶ ^ŝŐŶĂůeĚ ĐŽŶĐeŶƚƌĂtiŽŶ n=25 n=9n=22 n=13n=17 n=17n=12 n=12 τ .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint Figure 2. Single-molecule dynamics resolve Myosin binding/unbinding kinetics, revealing the origins of the ROCK/Myosin delay. (A-B) Near-TIRF microscopy image of single-molecule of Myosin, scale bar: 10 μm. (B) Colored circles: particles detected in image (A) by tracking, each molecule with a circle of a different color. (C) Time-lapse of the pulsed contraction in white box in (A). Upper right: number of detected particles. (D-G) Measurements from single- molecule microscopy data analysis of Myosin (NMY-2::GFP overexpression). N(embryos) = 5, N(pulses) = 38, thin line: median, thick line: average, shade: (D-F): std, (G): sem (D) Binding rate ( Kon) of Myosin as function of time, (E) unbinding rate ( koff), (F) normalized effective density, (G) signaled density ( Kon/koff, red) and effective density (blue) . (H) Variation of normalized values over time of ROCK intensity average (Fig. S1B), Myosin intensity average from Supp Fig. 1B, aligned and hence standing for av erage of Myosin intensity in Fig. 2F (red straight line), Myosin Kon average from Fig. 2D (cyan straight line), Myosin koff average from Fig. 2E (black straight line), and Myosin signaled density (Kon/koff) (magenta dashed line). (I-L) Single-molecule tracking of Myosin (NMY -2::GFP overexpression) in Myosin Phosphatase mutant context mel-11(it26). N(embryos) = 9, N(pulses) = 54, thin line: median, thick line: average, (I-K): std, (L): sem (I) Binding rate ( Kon), (J) Unbinding rate (K off), (K) Normalized effective density, (L) effective density (blue) and signaled density (Kon/Koff, red). (M) Descriptor to quantify the out -of-equilibrium dynamics of the Myosin, based on the ratio of signaled density amplitude over effective density amplitude for Myosin: (maximum of signaled density – min of signaled density)/(max of effective density – min of effective density) from (G) for the control and (L) for mel-11(it26), *** : p-value ≤ 0.001. .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint At equilibrium dN/dt = 0 “target value” kapp kapp N(t) * koff N(t) * koff N(t) = kapp/koff 0 30 60 90 0 0 0.0 0.5 1.0 0 0 10 20 30 number of moleculesbinding rate (s-1) unbinding rate (molecule-1.s-1) time (s) 100 200 300 0 0 100 200 300 0 0.0 0.5 1.0 0 0 10 20 30 number of particles time (s) unbinding rate (molecule-1.s-1)binding rate (s-1) A C At equilibrium dN/dt = 0 “target concentration” kapp kapp N(t) * koff N(t) * koff N(t) = kapp/koff D “infinite pool” kapp·dt N(t)·koff·dt N(t) dN = (kapp - koff· N)dt Effective value in numerical simulations Effective value in numerical simulations B time (s) ROCK normalized intensity 100 150 200 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 50 Figure 3 30 60 90 30 60 90 30 60 90 30 60 90 30 60 90 .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint Figure 3. Simple kinetic model supports kinetic control of the delay and highlights the role of unbinding kinetics in shaping Myosin accumulation . (A) Normalized intensity variation of ROCK as a function of time, over 6 successive pulsed contractions . (B) Schematic of the proposed model for the RhoA activation cascade. (C) Blue line: simulation of the effect RhoA periodic variation on the Kon (dark green line), with a constant koff (light green line), red line: Kon/koff. (D) Blue line: simulation of the effect RhoA periodic variation on the koff (light green line), with a constant Kon (dark green line), red line: Kon/koff. .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint *** *** UTR/NMY-2CYK-1/NMY-2 NMY-2/NMY-2CAP-1/NMY-2 *** ns time (s) 0 -5 5 10 15 20 -10 ~4.5s >0.5s ~4.5s RhoA Formin ROCK Myosin heavy-chainActin (proxy: GFP::AHPH) (CYK-1::GFP) (GFP::LET-502) (proxy: UTR::GFP) (NMY-2::mKate2) Myosin regulatory light-chain (MLC-4::GFP) >4.5s ns ns H I J K 0.0 0.5 1.0 1.5 2.0 -25 0 25 50 time (s) ďŝŶĚŝŶŐ ƌĂƚe (mŽůeĐƵůe͘s-1) 0.0 0.2 0.4 0.6 0.8 1.0 -25 0 25 50 time (s) ŶŽƌmĂůŝnjeĚ ƉƵůse ĐŽŶĐeŶƚƌĂtiŽŶ 0.00 0.05 0.10 0.15 -25 0 25 50 time (s) ƵŶďŝŶĚŝŶŐ ƌĂƚe (s-1) 0 10 20 30 40 50 -25 0 25 50 time (s) ƉƵůse ĐŽŶĐeŶƚƌĂtiŽŶ 0 1 2 3 4 0 -10 10 20 time (s) ďŝŶĚŝŶŐ ƌĂƚe (mŽůeĐƵůe͘s-1) 0.0 0.2 0.4 0.6 0.8 1.0 0 -10 10 20 time (s) ŶŽƌmĂůŝnjeĚ ƉƵůse ĐŽŶĐeŶƚƌĂtiŽŶ 0.00 0.05 0.10 0.15 0 -10 10 20 time (s) ƵŶďŝŶĚŝŶŐ ƌĂƚe (s-1) 0 10 20 30 40 50 60 70 0 -10 10 20 time (s) ƉƵůse ĐŽŶĐeŶƚƌĂtiŽŶ C D E F ^ŝŶŐůeͲmŽůeĐƵůe mLJŽsŝŶ unc-60(RNAi) ^ŝŶŐůeͲmŽůeĐƵůe ĂĐtiŶ A B 0.0 0.2 0.4 0.6 0.8 1.0 0 10 20 30 40 50 time(s) ŶŽƌmĂůŝnjeĚ ǀĂůƵes G kon ĂĐtiŶ koff ĂĐtiŶ ^ŝŐŶĂůeĚ ĐŽŶĐeŶƚƌĂtiŽŶ ŽĨ ĂĐtiŶ īeĐtiǀe ĐŽŶĐeŶƚƌĂtiŽŶ ŽĨ eůŽŶŐĂtiŶŐ ĨŽƌmŝŶ īeĐtiǀe ĐŽŶĐeŶƚƌĂtiŽŶ ŽĨ ĂĐtiŶ meĚŝĂŶ ĂǀeƌĂŐe meĚŝĂŶ ĂǀeƌĂŐe meĚŝĂŶ ĂǀeƌĂŐe meĚŝĂŶ ĂǀeƌĂŐe meĚŝĂŶ ĂǀeƌĂŐe meĚŝĂŶ ĂǀeƌĂŐe meĚŝĂŶ ĂǀeƌĂŐe meĚŝĂŶ ĂǀeƌĂŐe meĚŝĂŶ ĂǀeƌĂŐe meĚŝĂŶ ĂǀeƌĂŐe Figure 4 īeĐtiǀe ĐŽŶĐeŶƚƌĂtiŽŶ ^ŝŐŶĂůeĚ ĐŽŶĐeŶƚƌĂtiŽŶ īeĐtiǀe ĐŽŶĐeŶƚƌĂtiŽŶ ^ŝŐŶĂůeĚ ĐŽŶĐeŶƚƌĂtiŽŶ τ .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint Figure 4. Perturbing Actin dynamics does not alter Myosin unbinding kinetics . (A) Quantifications of time delay. Blue dots: time delay distri bution red over green channel, red dotted line: time delay of 0 s, blue shade: histogram of distribution. N(embryos) = 10, N(pulses) ≥ 200 (See Supplementary Table S1 and S2 for statistical details), ns: not significant, ***: p-value ≤ 0.001. (B) Summary of the observed time s delay within the RhoA activation cascade. (C-F) Single-molecule tracking of Myosin ::GFP in cofilin knockdown unc-60(RNAi). N(embryos) = 8, N(pulses) = 39. (C) Binding rate (Kon) of Myosin over time, (D) unbinding rate (koff), (E) normalized density, (F) effective density (blue) and signaled density (Kon/koff, red). (G- J) Single-molecule tracking of Actin ::GFP. N(embryos) = 6, N(pulses) = 55. (C-J) thin line: median, thick line: mean. (C-E, J-K) shade: std, (F, J) shade: sem. (G) Polymerization rate (Kon) of Actin over time, (H) unbinding rate (koff), (I) normalized density, (J) effective density (blue) and signaled density (Kon/koff, red). (K) Variation of normalized intensities over time of Actin average from Supp Fig. 1 C (green straight line), Formin super -diffusive population , aligned (red straight line), Actin Kon average from Fig. 4G (cyan straight line) , Actin koff average from Fig. 4H (black straight line), calculated signaled intensity value (Kon/koff) (magenta dashed line). .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint 0 A Figure 5 0 5 10 0.3 0.4 0.5 0.6Motor Contraction 0.2 0.1 0 0 5 10 0.05 0.1 0.15 0.2 Max. Actin Contraction 0 0 5 10 Delay (s) Sum of force [nN] D E B C Delay (s) 5 10 40 30 0 20 10 0 5 10 0 0.05 0.1 End-of-cycle Actin Contraction Delay (s) Delay (s) Delay (s) .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint Figure 5. Numerical simulations of actomyosin mechanics reveal that d elay between Actin assembly and Myosin accumulation affects force deployment during pulsed contractions. (A) Snapshots taken at the end of first myosin activation cycle with an increasing delay between formin -mediated Actin filament elongation and Myosin accumulation . (B) Quantification of the extent of motor contraction during pulsed contraction. (C) Quantification of actin contraction during pulsed contraction. (D) Sum of tensile forces acting and on formin- elongated filaments. (E) Average force generated locally and acting on the overall network. .CC-BY-NC 4.0 International licenseperpetuity. It is made available under a preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in The copyright holder for thisthis version posted February 10, 2026. ; https://doi.org/10.64898/2026.02.08.704688doi: bioRxiv preprint