A self-organized synthetic morphogenic liposome responds with shape changes to local light cues

The study asks how localized extracellular signals (morphogens) can be transduced into cytoskeletal reorganizations that drive cell morphogenesis, and how prior morphological states affect responsiveness. In cells, Rho GTPases (e.g., Rac) recruit kinases (e.g., PAK1) to membranes (dimensionality reduction), phosphorylating stathmin to relieve its inhibition of microtubule (MT) growth, thereby steering MTs toward stimuli. The authors note that cells integrate memory of prior cues to commit to shapes, but mechanisms of this memory in morphogenesis are unclear. To dissect minimal principles, they reconstitute a light-responsive system in giant unilamellar vesicles (GUVs) containing dynamic MT-asters and a stathmin phosphorylation cycle, enabling analysis of coupling between signaling gradients and MT-induced membrane deformations and how initial shapes influence responses.

Background emphasizes: (1) Cytoskeletal systems (actin for rapid peripheral dynamics; MTs for long-lived, global organization) shape cells and respond to extracellular morphogens. (2) Rho GTPases recruit kinases (e.g., PAK1) to membranes, increasing local activity via dimensionality reduction; kinases phosphorylate stathmin, a tubulin-sequestering protein, to promote MT growth. (3) Prior work established stathmin-tubulin interaction gradients in cells and MT dynamic instability, as well as effects of confinement on MT organization, and concept of stigmergy in pattern formation. (4) Open question: how memory of prior morphogen patterns is maintained and affects morphogenesis. The paper builds on these by reconstituting a minimal signaling-aster system to probe self-organization and stimulus responses.

• Encapsulation of MT-asters: Purified centrosomes, tubulin (with 10% fluorescently labeled tubulin), and GTP were encapsulated into ~25±5 µm GUVs using cDICE. GUV membrane tension was controlled via osmolarity (iso-osmotic rigid vs hyperosmotic deformable). Aster size controlled by tubulin concentration (15–40 µM). Imaging by confocal laser scanning microscopy (CLSM) at 33–34 °C.
• Morphometrics: Quantified centrosome position (0 center to 1 membrane-proximal), GUV eccentricity, average MT length, angular membrane curvature kymographs, and Relative Centrosome Displacement (RCD). Defined Relative Centrosome Surface (RCS) to estimate MT nucleation density.
• Temperature ramp: Raised from ~21 °C to 34 °C to induce net MT growth to probe transitions under uniform cytoplasmic-like signals.
• Modeling MT–membrane subsystem: Agent-based Monte Carlo simulations (1D circular geometry) incorporating dynamic instability and self-induced capture (SIC) of MTs into protrusions; complementary reaction–diffusion (RD) substrate-depletion model for bundled vs free MTs.
• Stathmin regulation assays: Single-filament TIRF microscopy measured MT growth velocity and catastrophe frequency vs tubulin concentration and stathmin/phospho-stathmin (pStathmin) fractions. Aster size on glass measured vs stathmin/pStathmin.
• Light-inducible signaling: Reconstituted C2-iLID (membrane-targeted) and SspB-AuroraB kinase fusions (plus SspB-phosphatase lambda, PPX/PPA) to control membrane recruitment by 488 nm light. Quantified dose-dependent recruitment kinetics and spatial gradients upon local irradiation; examined clustering via membrane fluorescence mapping and recurrence-based information entropy analysis.
• Self-organization of signaling: Established cooperative clustering (CC) model for SspB-AuroraB monomers→clusters with monomer depletion; simulated pattern formation on 2D grids and analyzed entropy changes; compared to SspB-PPX control.
• Stathmin phosphorylation cycle and gradients: Developed COPY FRET sensor variants to measure stathmin phosphorylation in bulk and inside GUVs using FLIM-FRET and ratiometric imaging. Measured steady-state phosphorylation with kinase and phosphatase present; modeled reaction-diffusion in radial 1D geometry using measured kinetic parameters to predict pStathmin and free tubulin gradients (decay length ~submicron).
• SynMMS construction: Co-encapsulated tubulin647 (40–44 µM total, 10% labeled), stathmin (~4–5 µM), C2-iLID (~5 µM), SspB-AuroraB (~4.9–12 µM), PPX (~0.5 µM), centrosomes, ATP/GTP. Imaged responses to global or local 488 nm irradiation under rigid or deformable membranes. Compared to controls lacking stathmin (SynMMS-stat).
• Coupled-system theory: Built an RD model coupling SIC (MT bundling) and CC (AuroraB clustering), with intersystem links: AuroraB clusters promote tubulin gradient → MT growth; MT-induced membrane deformation enhances AuroraB clustering. Linear perturbation analysis and simulations explored regimes (SIC>CC vs CC>SIC), global/local stimuli, and state transitions.
• Data analysis: Image processing to enhance MT bundles; quantification of membrane protein translocation correcting for luminal contributions; kymographs; recurrence entropy metrics; statistics via Kolmogorov–Smirnov tests.

• MT–membrane subsystem self-organization: High tubulin (35–40 µM) in rigid GUVs produced cortical MTs and decentered centrosomes; in deformable GUVs, MTs generated spiking protrusions (SPs) with bundled MTs converging into protrusions and increased eccentricity and MT length, consistent with self-induced capture (SIC). Temperature-induced net MT growth drove transitions to polar protrusions via protrusion coalescence and centrosome decentering.
• Modeling confirmed SIC-driven polar patterns; increased MT nucleation favored star-like multi-protrusion patterns, captured by both Monte Carlo and RD substrate-depletion frameworks.
• Stathmin regulation: Increasing stathmin decreased MT growth speed and increased catastrophe frequency; phosphorylation reversed these effects. On surfaces, stathmin reduced aster size, pStathmin had minimal effect. In GUVs (40±7 µM tubulin), 5±1 µM stathmin yielded spherical morphologies with shorter MTs; 5±1 µM pStathmin restored polar morphology akin to high tubulin alone. Fig. 2d reported highly significant differences (e.g., p-values ~5.5×10⁻10 to 4.2×10⁻10 for centrosome position, eccentricity, MT length).
• Light-inducible kinase recruitment: 488 nm irradiation caused rapid (∼10 s) localized SspB-AuroraB membrane recruitment with gradients reaching steady state in ∼50 s; gradients for SspB-PPX were shallower. AuroraB formed small clusters upon recruitment (irrespective of ATP), unlike PPX; cluster pattern regularity increased with membrane recruitment (lower information entropy upon randomization), indicating cooperative clustering (CC) with monomer depletion.
• Stathmin phosphorylation/tubulin gradients: In solution, combined SspB-AuroraB and PPX maintained steady-state pStathmin (~20%) consistent with kinetics (PPA kcat/KM ≈ 2.2×10⁴ s⁻¹ M⁻¹; kinase ≈ 1.1×10³ s⁻¹ M⁻¹). In GUVs with PPX, light-induced SspB-AuroraB recruitment created a steep pStathmin gradient emanating from the membrane (~0.5 µm decay length) and, via tubulin release from pStathmin, a membrane-proximal free-tubulin gradient maintained by a phosphorylation/dephosphorylation cycle and diffusion.
• SynMMS responses: In rigid GUVs, light-induced recruitment increased RCD (centrosome decentering) reversible upon light removal; effect required stathmin (no change in SynMMS-stat). In deflated SynMMS with polar protrusions, global light elongated and reoriented liposomes; SspB-AuroraB accumulated in protrusions over time.
• Initial morphologies: SynMMS displayed spherical, polar, and star-like states. Star-like morphologies (axially distributed protrusions) were more frequent with signaling present than asters alone, suggesting basal AuroraB signaling stabilizes protrusions. Protrusion types included SPs and membrane-sheet protrusions (MSPs); MSPs occurred only with signaling. Astral-MT density (RCS) increased from spherical to polar to star.
• De novo protrusion formation: Strong global light in spherical SynMMS with sparse asters induced multiple stable MSPs that accumulated SspB-AuroraB and coupled to centrosome movement; absent in SynMMS-stat. Local irradiation sequentially induced MSPs/SPs in targeted regions, demonstrating controlled de novo morphogenesis dependent on stathmin.
• Reciprocal coupling: Preformed protrusions preferentially recruited SspB-AuroraB at low light doses, enhancing local MT growth and pushing centrosomes to the periphery; star-like SynMMS transitioned to spherical/cortical bundles upon strong activation, whereas polar SynMMS broadened main protrusions. Feedback arose from geometry-facilitated recruitment, slowed lateral diffusion within protrusions, and limited local stathmin/PPX increasing tubulin gradient amplitude.
• Coupled-system theory: SIC-dominant (γ1 high) yielded polar states robust to distal local stimuli but reorientable when stimuli are proximal. CC-dominant (γ2 high) stabilized star-like patterns robust to global stimulation; local stimuli caused coalescence into a polar bundle oriented toward the cue. Thus, initial states determine responsiveness and directionality.
• Light-guided morphogenesis depends on initial state: In star-like states with strong CC (high translocation), local irradiation drove convergence of protrusions and a star-to-polar transition toward the stimulus with explosive sprouting in the stimulated region. With low translocation (weak CC), fluctuating SPs redistributed anisotropically toward the signal without stabilization. Polar states rotated toward proximal cues but were robust against distal cues, with induced MSPs sliding and being captured by the main protrusion.

The reconstituted SynMMS demonstrates that localized signal transduction via dimensionality reduction can generate a membrane-proximal stathmin phosphorylation and free tubulin gradient sufficient to bias MT growth and drive membrane deformations. Reciprocal coupling arises because MT-induced deformations enhance membrane recruitment and clustering of kinase, reinforcing local signaling and stabilizing protrusions. Each subsystem (MT SIC and AuroraB CC) independently exhibits self-organization via self-amplification with substrate depletion; their coupling yields emergent morphologies whose plasticity and directional responses depend on the balance between SIC and CC. Star-like states (CC>SIC) are highly plastic and can reorient globally toward local cues via coalescence, whereas polar states (SIC>CC) are robust, responding only to proximal cues by reorientation without morphological transitions. These findings provide mechanistic insight into how cells might integrate prior morphological states (memory) with external morphogens to control morphogenesis, consistent with stigmergic principles of matter coalescence.

This work establishes a minimal, light-responsive synthetic morphogenic system in which a stathmin phosphorylation cycle couples signaling to MT-driven membrane remodeling. Key contributions include: (1) demonstration of a dynamically maintained membrane-proximal pStathmin and free-tubulin gradient upon localized kinase recruitment; (2) evidence for cooperative clustering of AuroraB on membranes and SIC-driven MT bundling as dual self-organizing modules; (3) elucidation of reciprocal coupling between signaling and MT-induced deformations that yields global shape changes oriented by local cues; (4) a theoretical framework linking initial morphological states (SIC vs CC dominance) to robustness and plasticity of responses. Future directions include improving control over initial conditions (e.g., defined nucleators and protein insertion methods), extending lifetime via energy regeneration (photosynthetic ATP/GTP production), integrating actin networks for fast local dynamics, and advancing toward self-replicating and communicating synthetic tissues.

• Variability in initial morphologies due to: (1) variable encapsulation efficiency of C2-iLID affecting basal and maximal signaling (thus CC strength); (2) variable MT nucleation on centrosome clusters affecting astral MT density. These limit precise control over starting states.
• Short operational lifetime constrained by finite encapsulated ATP/GTP fuel.
• Some lipid irregularities can cause inactive aggregates; care needed to exclude such areas in analyses.
• Models simplify complex biochemistry (e.g., stathmin multi-site phosphorylation) and MT stochasticity; qualitative but not fully quantitative predictions.
• Observation constraints (confocal plane) may miss out-of-focus protrusions, necessitating indirect curvature/transmission tracking.