Photoreceptor‐Like Signal Transduction Between Polymer‐Based Protocells

Abstract

Deciphering inter‐ and intracellular signaling pathways is pivotal for understanding the intricate communication networks that orchestrate life's dynamics. Communication models involving bottom‐up construction of protocells are emerging but often lack specialized compartments sufficiently robust and hierarchically organized to perform spatiotemporally defined signaling. Here, the modular construction of communicating polymer‐based protocells designed to mimic the transduction of information in retinal photoreceptors is presented. Microfluidics is used to generate polymeric protocells subcompartmentalized by specialized artificial organelles. In one protocell population, light triggers artificial organelles with membrane‐embedded photoresponsive rotary molecular motors to set off a sequence of reactions starting with the release of encapsulated signaling molecules into the lumen. Intercellular communication is mediated by signal transfer across membranes to protocells containing catalytic artificial organelles as subcompartments, whose signal conversion can be modulated by environmental calcium. Signal propagation also requires selective permeability of the diverse compartments. By segregating artificial organelles in distinct protocells, a sequential chain of reactions mediating intercellular communication is created that is further modulated by adding extracellular messengers. This connective behavior offers the potential for a deeper understanding of signaling pathways and faster integration of proto‐ and living cells, with the unique advantage of controlling each step by bio‐relevant signals.

This study presents polymer‐based protocells containing light‐sensitive and catalytic artificial organelles (AOs), forming a communicating protocellular system. Light‐triggered substrate release from AOs in sender protocells is used to relay a signal to a receiver protocell where its conversion takes place in catalytic AOs. This system, modulated by environmental calcium, can enhance the understanding of cellular communication networks.

1. Introduction

Living cells receive, transmit and process signals from their environment via a plethora of inter‐ and intracellular signaling pathways.^{[ 1} , ^{2 ]} In multicellular communities, sensing the cellular environment and exchanging information between cells is prerequisite for coordinated behavior. To gain a holistic understanding of cellular communities and coordination, it is essential to explore underlying principles of inter‐ and intracellular communication. Models of communication networks have been instrumental in elucidating the coordination of behavior among cells.^{[ 3} , ⁴ , ^{5 ]} The bottom‐up design of synthetic cellular models provides the means to study and understand complex cellular processes in simplified systems.^{[ 6 ]} Giant unilamellar vesicles (GUVs, 1–100 µm in diameter) bound by membranes assembled from amphiphilic molecules such as lipids, peptides or amphiphilic block copolymers lend themselves to mimicking natural cells both in size and membrane structure.^{[ 6} , ^{7 ]} Several fundamental cellular processes such as protein production,^{[ 8 ]} or DNA‐ and self‐replication,^{[ 9} , ^{10 ]} have been successfully mimicked in protocells. While such reductive systems may contribute to a better understanding of simple cellular processes, complex intracellular processes are impeded by the absence of cell‐mimetic organization and structures, i.e., organelles. By compartmentalizing protocells through the encapsulation of artificial organelles (AOs), natural processes can be more closely modeled.^{[ 11} , ¹² , ^{13 ]} Over the past few years, bottom‐up designed synthetic cell mimics have also evolved as models for either intra‐ or intercellular communication.^{[ 14} , ¹⁵ , ^{16 ]} The majority of these studies are based on lipidic micrometer‐sized compartments, each typically containing a single type of subcompartment, if any.^{[ 3} , ¹¹ , ¹² , ¹⁷ , ¹⁸ , ^{19 ]} For example, a vesicle‐in‐vesicle structure constructed from lipids was shown to secrete insulin contained in the smaller vesicles by fusion with the larger vesicle membrane, triggered by a decrease in pH inside the larger vesicle in response to the enzymatic degradation of externally added glucose.^{[ 20 ]} Communication between lipidic giant unilamellar vesicles was achieved by a bioluminescence signal produced intracellularly in GUVs exposing blue light responsive protein, iLID, on their surface, which then mediated binding to GUVs exposing Nano protein under blue light.^{[ 19 ]} Upon binding, contact dependent transfer of secondary signals between the GUVs occurred. However, lipid‐based synthetic systems are often mechanically unstable and prone to fusion and rupture due to osmotic imbalances. Their delicate nature not only renders them sensitive to changes in temperature, pH and other environmental factors but can make their analysis technically demanding. GUVs arising from the self‐assembly of polymers, on the other hand, offer greater stability and versatility in membrane properties due the wide range of polymer chemistries. Another advantage of polymer GUVs and nanosized polymer‐based compartments is that their membrane building blocks can be chemically produced to obtain significantly less permeable membranes than those of lipid vesicles, allowing to control molecular flow to and from the compartments by insertion of suitable communication channels such as membrane pores,^{[ 21} , ²² , ^{23 ]} membrane proteins,^{[ 24} , ²⁵ , ^{26 ]} or pore‐forming peptides.^{[ 27} , ²⁸ , ^{29 ]} One prominent gap in the field is being able to more closely emulate the complex sequential reactions of natural cells, involving subcompartments from different cell populations. The inability to replicate controlled, successive reactions across multiple membrane boundaries in response to environmental signals, as observed in collective cellular behaviors, represents a major challenge not yet addressed in a synthetic system. Here, we follow a bottom‐up strategy to construct specialized polymer‐based protocells and integrate an organelle‐based chain of reactions into intracellular communication in response to environmental signals (Figure  1 ). Engineered sequential signaling reactions involve molecular flow between various membrane barriers (organelle and protocell membranes, and across boundaries of different protocells), similarly to natural cells. Therefore, they require precise control over i) the subcompartmentalization architecture, ii) the stability of the protocells, and iii) the timing of a specific sequence to promote overall functionality in response to stimuli. In vision, light signals are transduced in vision through rhodopsin molecules located in the membranes of outer segments of photoreceptor cells, initiating a complex signaling cascade that ultimately leads to neurotransmitter release for visual processing.^{[ 30 ]} By constructing a compartmentalized sender protocell in which the release of a signaling molecule is triggered by light activation of synthetic rotary motors in the membrane of artificial organelles (AOs),^{[ 31 ]} we emulate a photoreceptive cell. Although lipid‐based nanovesicles embedded with photoreceptors were previously reported to show specific color sensitivity, they were not further exploited in the context of an artificial signaling model addressing communication.^{[ 32 ]} Our protocells and artificial organelles are based on an amphiphilic diblock copolymer, poly(dimethylsiloxane)‐block‐poly(2‐methyl‐2‐oxazoline) (PDMS‐b‐PMOXA), which is known to assemble highly stable nano‐ and micrometer‐sized vesicles that are able to encapsulate a variety of molecules in their aqueous cavity.^{[ 7} , ³³ , ^{34 ]} In addition, membranes of PDMS‐b‐PMOXA based vesicles have the necessary fluidity for the incorporation of native and artificial membrane constituents, resulting in increased functionality, permeability and stimulant sensitivity of the vesicles.^{[ 31} , ³⁵ , ³⁶ , ³⁷ , ^{38 ]} Stimuli‐responsive systems have emerged as promising tools for directing the physical transformations of membranes in response to external stimuli^{[ 34} , ³⁹ , ^{40 ]} and among these stimuli, light offers exceptionally precise spatiotemporal control.^{[ 31} , ^{41 ]} In particular, we have recently presented a PDMS‐b‐PMOXA‐based polymersome system with membrane‐embedded synthetic rotary molecular motors for light‐activated drug delivery.^{[ 31 ]} However, we did not explore their potential to subcompartmenalize protocells and function as artificial organelles in a signaling sequence. Light‐driven synthetic rotary motors, derived from chiroptical molecular switches, exhibit a higher degree of controlled motion. First reported by Feringa et al. in 1999,^{[ 42 ]} they represented a breakthrough in the field of molecular motors due to their repetitive, photochemically driven unidirectional rotation around a central carbon–carbon double bond. Extensive reviews about the design, applications and function of these molecular motors can be found elsewhere.^{[ 43} , ^{44 ]} Besides the molecular motor‐based AOs, we generated catalytic AOs harboring an enzyme inside and membrane‐embedded biopores that rendered the membrane permeable for substrate and product diffusion. Catalytic AOs enabled the construction of subcompartmentalized receiver protocells. AOs of defined size (≈0.1 µm) that mimic naturally occurring organelles were created through film rehydration and subsequently extruded.^{[ 38 ]} For the production of AO‐containing protocells, a high‐throughput microfluidic setup was used because of the ease of cargo encapsulation and extraordinary reproducibility, essential for establishing a modular system to mimic defined aspects of natural cells.^{[ 29} , ⁴⁵ , ⁴⁶ , ⁴⁷ , ⁴⁸ , ^{49 ]} Furthermore, using the selected block copolymer in microfluidic formation enables our system to overcome issues of poor encapsulation efficiency and unstable, leaky vesicles.^{[ 5} , ^{12 ]}

Figure 1.

Schematic of light‐triggered, organelle‐based chain of reactions mediating intercellular communication between sender and receiver protocells, each with a distinct set of functional AOs. One type of AO contains light‐responsive rotary molecular motors in their membranes, enabling the controlled release of “signaling” molecules in sender protocells upon illumination. Signaling molecules are transduced through membrane pores in both sender and receiver protocells to the second type of AO with catalytic function inside receiver protocells, where they act as substrates that are converted into reporting molecules. Signal transduction in receiver protocells can be modulated by the addition of environmental CaCl₂.

Rod and cone cells below the pigment epithelium of the eye perform a series of signal transduction reactions upon light exposure.^{[ 50 ]} To elicit a light‐triggered response in a synthetic system, we utilized photoresponsive polymersome‐derived AOs to initiate a cascade reaction by releasing a substrate,^{[ 31 ]} which can subsequently be processed downstream by catalytic AOs. Progressing toward the creation of an artificial retinal synapse,^{[ 51 ]} we established an organelle‐based chain of reactions mediating intercellular communication by spatially sequestering substrate‐releasing and catalytic AOs in sender and receiver protocells. While light initiates the signal transduction in the retina, Ca²⁺ in the outer segment controls photoreceptor light adaptation and at the synaptic terminal is an important modulator of downstream signaling.^{[ 52 ]} Likewise, we demonstrate dual sensitivity to Ca²⁺ and light by co‐encapsulating photoresponsive and calcium‐responsive AOs within the same protocell. More importantly, changes in the environmental calcium levels enable the modulation of the downstream signaling response in corresponding receiver protocells. We believe our system represents a unique prototype for the bottom‐up engineering of an artificial cellular system with connective behavior by combining advanced polymer materials, biomolecules, synthetic molecular motors, and high‐throughput microfluidic technology, thereby helping to expand the fundamental understanding of signal transduction in biological systems.^{[ 53 ]} We had to overcome substantial challenges for a successful combination without: i) affecting the building blocks' integrity, ii) aggregation, iii) loss of functionality, iv) losing of the desired direction of the sequential reaction, etc. Overcoming these hurdles to arrive at a successful sequential organelle‐based chain of reactions that lay the grounds for intercellular communication opens new avenues in the field of bottom‐up construction of synthetic cells.

2. Constructing Artificial Organelles and Protocells with Block Copolymer Membranes

To mimic nanometer‐sized natural organelles, different types of AOs, characterized by specific functionality, were engineered. The design of AOs was based on the fine‐tuning and optimization of self‐assembled polymersomes as nanocompartments, each specifically enriched with components that provide key functionalities for our protocells. In brief, AOs were formed by rehydrating a thin layer of PDMS₂₅‐b‐PMOXA₁₀ copolymer overnight with a buffer containing the desired bioactive molecules (see corresponding sections for details). Subsequent extrusion reduced the diameter of the resulting AOs to ≈200 nm as determined by dynamic/static light scattering (DLS/SLS) and nanoparticle tracking analysis (NTA) (Table S1, Supporting Information). The radius of gyration (R _g) was calculated by SLS and the hydrodynamic radius (R _h) using the DLS profile. The shape parameter ρ (R _g/R _h) of ≈1 is typical for the morphology of hollow spheres (Table S1, Supporting Information).^{[ 54} , ^{55 ]}

Through a one‐step microfluidic droplet formation process using a six‐way junction on a microfluidic silicon‐glass chip (Figure S1, Supporting Information), double emulsions serving as templates for polymer GUVs were produced at high throughput.^{[ 29} , ^{45 ]} Specifically, an inner aqueous phase (200 mm sucrose in PBS (467 ± 1.1 mOsmol kg⁻¹)) and an enclosing polymer‐containing organic phase consisting of PDMS₂₅‐b‐PMOXA₁₀ block copolymer in a mixture of hexane and chloroform (3:2 v/v) were flowed together, while an outer aqueous phase (100 mm NaCl, 5% PEG₃₅₀₀₀, 0.1% Pluronics F‐68 in PBS (549 ± 2.3 mOsmol kg⁻¹)) was cross‐flowed to break the outer aqueous phase and polymer‐containing organic phase flow into monodisperse double emulsion droplets. The resulting double emulsion droplets were exposed to air where evaporation of the volatile organic solvent led to the formation of GUVs with an average diameter of 43.2 ± 0.7 µm (± 1.6%, Figure S2, Supporting Information) within minutes.^{[ 29} , ^{45 ]} Evaluation of GUV stability revealed 80% intact GUVs after 4 days and up to 40% after 1 month (Figure S3a,b, Supporting Information). This stability is significantly higher than that of lipidic GUVs, which typically last for only a few hours.^{[ 56} , ^{57 ]}

3. Integrating AOs in GUVs for Advanced Compartmentalization of Protocells

The use of microfluidics in the construction of hierarchical membrane‐bound protocells offers unparalleled precision and control over their membrane composition and luminal contents that is fundamental for replicating complex cellular structures and functions. For the creation of a hierarchical protocell capable of signaling, it is essential to ensure controlled encapsulation of AOs in GUVs without compromising their functionality (Figure  2 ). To assess and optimize AO encapsulation in GUVs, we prepared corresponding polymersomes with ATTO488 fluorescent dye as cargo and used them as AOs that can be quantitatively assessed based on their fluorescence (AO_A488, Table S1, Figure S4, Supporting Information). These AOs were added directly to the inner aqueous phase at concentrations of 3.4 × 10¹¹–3.4 × 10⁹ AO mL⁻¹ (Figure 2a). The composition of inner and outer aqueous phase was optimized to obtain stable GUVs while not compromising the stability or function of the nano‐sized compartments to be encapsulated within the GUVs (Figure S3b, Supporting Information). Concentration‐dependent encapsulation into GUVs was evaluated based on fluorescence microscopy images (Figure 2b,c) and no impairment of microfluidic GUV generation was observed even at the highest AO concentration (3.4 × 10¹¹ AOs mL⁻¹, Figure S5, Supporting Information). The number of encapsulated AOs per protocell was determined based on 3D reconstructions of Z‐stacked images recorded by confocal fluorescence microscopy (Figure 2d) and compared to theoretical loading values calculated from AO input concentrations and protocell volumes (Figure 2e; Table S2, Supporting Information). A clear linear correlation (r² = 0.93) was found between the AO input concentration and the observed loading at an encapsulation efficiency of 110.2 ± 20.7% (± 18.2%, Figure S6, Supporting Information). Thus, microfluidics enables precise control over the number of encapsulated AOs per protocell with unprecedented encapsulation efficiency and homogeneity of GUVs, a level of control not achievable through film rehydration.^{[ 11} , ^{12 ]}

Figure 2.

Formation and characterization of compartmentalized protocells. a) Bioactive cargo molecules are encapsulated in self‐assembled artificial organelles by film rehydration, forming functional AOs. Subsequently, stimuli‐responsive AOs are encapsulated into GUVs using double emulsion microfluidics. b,c) Representative fluorescence micrographs of ATTO488‐loaded AOs (green) in protocells (cyan) at inner aqueous phase concentrations of b) 3.4 × 10¹⁰ AO and c) 3.4 × 10¹¹ AO mL⁻¹. Scale bars, 10 µm. d) 3d reconstruction of AO_A488 distribution (green) within a protocell stained with BODIPY 630/650 (red). e) Experimentally determined loading numbers of AO models per protocell (black) from 3D reconstructed protocells (n > 3) and theoretical loading values (blue) determined by input AO concentration and protocell volume (n = 3). Data is expressed as a mean ± SD. f) 3D AO distribution within a 2 µm segment of a protocell (concentration of 3.4 × 10¹⁰ AO mL⁻¹) visualized through a 2D Kernel Density Estimation plot with 20 levels along the x and y axes. Histograms represent the AO density distribution along the x and y axes. The symmetric and broad density contours suggest a uniform particle distribution in the sphere's volume, with a higher concentration at the core (n = 3). g) FCS autocorrelation curves for AO_A488 encapsulated in the cavity of protocells at different concentrations (n = 30). Dotted lines represent raw data averages and solid lines represent fitted curves.

The 3D protocell reconstructions revealed that the encapsulated AO_A488 retained their structural integrity, were evenly distributed throughout the protocell (Figure 2f; Figures S5 and S7, Supporting Information) without detectable aggregation or membrane interaction, even at the highest AO concentration (3.4 × 10¹¹ AO mL⁻¹, Figure S8, Supporting Information). Analysis of the average nearest neighbor distance between AOs revealed distances between 1 and 3.3 µm that is in agreement with inter‐organelle distances found between lysosomes, a storage site for Ca²⁺ in natural cells.^{[ 58 ]} We optimized each type of protocell according to two key requirements: i) ensuring the distances between the respective AOs are bio‐relevant, and ii) selecting AO concentrations at a given number of protocells to promote an efficient cascade. The inter‐AO distance can be modeled using an ideal, dense packing of AOs in the protocell in a simple cubic structure, corroborating a dense, uniform distribution of AOs within the protocells (Figure S9, Supporting Information). Furthermore, no difference was detected between the size of protocells (Figure S10, Table S3, Supporting Information) or the motion of AOs at different concentrations in protocells, suggesting that the protocell volume represents a confined space, but does not affect the motion of AOs (Movies S1 and S2, Supporting Information). Further investigations on the motion of AO_A488 in solution and in protocells were performed using fluorescence correlation spectroscopy (FCS) (Figure 2f; Figure S11, Supporting Information). Comparison of the diffusion times (τ _D) of AO_A488 before and after encapsulation confirmed that neither the buffer composition nor the confinement affected AO diffusion, and no AO aggregation was observed inside protocells (Figure S11, Supporting Information). The increasing correlation G(τ) was a consequence of the number of AO_A488 in the cavity of protocells and inversely proportional to their concentration (Figure 2g). FCS measurements in the extravesicular environment of the protocells showed no diffusion times corresponding to AO_A488, indicating the high stability of protocells as no AOs were released from the protocells (Figures S11 and S12, Supporting Information).

4. Protocells Responding to Light

The creation of protocell communication hinges on the controlled release of encapsulated compounds from AOs. To achieve this, we first established and optimized light‐mediated cargo release from stimulus‐responsive AOs, using ATTO488 encapsulated in AOs with a synthetic dibromo light‐driven molecular motor incorporated in the hydrophobic part of their membrane (AO_MM_A488, Figure 3a, Table S1, Figure S13, Supporting Information).^{[ 31 ]} The molecular motor is able to rotate unidirectionally across the central carbon‐carbon double bond when irradiated with light at ≈430 nm.^{[ 31 ]} It has recently been shown that light‐induced molecular motor rotation leads to membrane rupture and a release of drug from the AO cavity.^{[ 31 ]} Symmetric molecular motors undergo a 360 degree rotation cycle involving two equivalent photochemical E‐Z isomerizations, each followed by a thermal helix inversion step, resulting in continuous unidirectional motion along the central axis. As in the current study, MM‐polymersomes are intended to function as membrane‐bound organelles for temporary storage of substrate molecules and a rapid release “on demand” inside synthetic protocells, they required extensive modifications. The intrinsic properties of the specific cargo demanded to adapt the self‐assembly process of MM‐AO formation and to optimize the illumination time to achieve rapid substrate release inside protocells. Moreover, modifying the number of molecular motors per polymersome integrated in the membrane, allows to tune membrane permeability or entirely disintegrate it with temporal precision. In addition, as stability and light‐responsiveness are affected by the composition of the protocell interior, illumination conditions had to be adapted for MM‐AOs residing inside protocells. The successful insertion and rotational performance of the molecular motors into our AOs was demonstrated by UV–vis spectroscopy of empty AO_MM (Figure S14, Supporting Information). The spectral changes observed for the molecular motor peak in a dodecane solution (λ_max = 405 nm, Figure S14a,b, Supporting Information) and embedded in the AO membrane (Figure S14c,d, Supporting Information) confirmed the successful photochemical E‐Z isomerization and thus the rotation of the molecular motor within our system, resulting in the rupture of AOs (Figure S15, Supporting Information). Studies on the rotation cycle of the molecular motor showed a decrease in the photoisomerization quantum yield of around half, compared to the free molecular motor in solution, while the rotation speed of the integrated molecular motor increased 1.4‐fold due to the shielding effect of the hydrophobic AO membrane domain (Figure S16, Table S4, Supporting Information).^{[ 41 ]} As seen in previous studies with similar molecular motors embedded in lipid bilayers, a decrease in the quantum yield could be explained by the tendency of the molecular motors to form small aggregates due to enhanced π−π interactions in hydrophobic environments.^{[ 41 ]} This aggregation state may hinder the isomerization step, which involves a large conformational change, leading to the observed reduction in quantum yield. On the other hand, an increase in rotation speed is attributed to a higher preference of the motor to undergo thermal helix inversion favoring intermolecular interactions in ordered environments. The release of cargo was evaluated by the release of ATTO488 from AO_MM_A488 in response to irradiation (λ = 430 nm) for up to 20 min. A ≈15‐fold increase in fluorescence intensity together with a decreased dye release rate ($\frac{d}{dx}$ of slope) was observed after 10 minutes of irradiation for free AO_MM_488 (Figure 3a; Figure S17a,b, Supporting Information). AOs without motors (AO_A488) were stable regardless of the irradiation conditions and did not release their cargo (Figure 3a; Figure S17c, Supporting Information).^{[ 31 ]} While a similar release rate plateau after ≈10 min of light exposure was observed for encapsulated AO_MM_488 in protocells (Figure S18, Supporting Information), the relative fluorescence increase was lower compared to free AO_MM_488 (Figure 3b) due to the lower relative concentration per protocell (≈6.7 ± 1.2 × 10⁹ AO protocell⁻¹). In addition, the membrane of the protocells and the complexity inside induce a decrease in the light intensity that will reduce the release efficacy of AOs. While the presence of molecular motors in AOs in protocells had little destabilizing effect on the protocells with up to 90% intact protocells after 1 day, light‐induced motor activation reduced the protocell's stability after 1 day (Figure S19, Supporting Information), resulting in slightly reduced long‐term stability. The stability of the protocells was optimized by adjusting the number of AOs per protocell with high release efficiency. However, this optimization involved a tradeoff, resulting in a slightly lower long‐term stability. The stability could be further increased by i) decreasing the number of AO_MM per protocell, ii) decreasing the number of molecular motors per AO, or iii) cross‐linking the polymer membrane of the protocells. Consequently, we always added freshly prepared AO_MM to minimize destabilization. Furthermore, the presence of MM_AOs in protocells had no significant influence on the polymer membrane permeability, even upon illumination of the MM_AOs (Figure S20, Supporting Information).

Figure 3.

Light‐triggered intracellular signaling between AOs. a) Schematic representation of stimuli responsive AOs containing molecular motors (MM) in their membranes. Upon illumination with light (hv) at 430 nm, the molecular motors rotate, thereby destabilizing the polymer membrane and releasing the cargo from the AO lumen. Release of ATTO488 dye from the AO lumen upon exposure at 430 nm. Fluorescence measured from fluorescence micrographs (n = 3). b) Release of ATTO488 dye from the intracellular AO upon exposure at 430 nm for up to 20 min with 2 min intervals. Fluorescence measured in the lumen of protocells from fluorescence micrographs (n = 3). c) Schematic representation of stimuli‐responsive intracellular signaling cascade between AOs in a protocell. Upon irradiation, MM‐AO release FDG that can diffuse through melittin pores into a second, βGal‐encapsulating AO. The βGal hydrolyzes the non‐fluorescent substrate FDG to the fluorescent fluorescein product. Normalized fluorescence inside protocells encapsulating AO_MM_FDG and AO_βGal with and without melittin pores after 1 h of incubation (n = 5). Significance levels: p > 0.05 (n.s.), p < 0.05 (*), p < 0.005 (**), and p < 0.0005 (***).

5. Engineering Artificial Intracellular Signaling Pathways

We then proceeded to develop a two‐compartment intracellular signaling cascade inside a single photoreceptor protocell to study the intracellular signal propagation and more closely replicate the complexity of natural signaling cascades in vitro.^{[ 59 ]} Specifically, we aimed to mimic intracellular signaling in response to external stimuli, using precisely designed AOs (AO_MM_FDG, Table S1, Figure S21, Supporting Information) that released the non‐fluorescent substrate fluorescein‐di‐β‐D‐galactopyranoside (FDG) upon light activation. Subsequently, the FDG was converted to fluorescein by β‐galactosidase (βGal) within melittin‐permeabilized AOs present inside the same protocells (AO_mel_βGal, Figure 3c; Table S1, Figure S22, Supporting Information). The number of melittin pores per AO_mel_βGal was calculated to be 177 ± 37 by FCS of Cy5‐melittin‐incorporating AOs (Figure S23a, Supporting Information). The encapsulation efficiency of βGal within the AOs’ cavities was estimated at 31% ± 17% by the bicinchoninic acid assay (Figure S23b, Supporting Information).^{[ 60 ]}

In a first bulk experiment comparing the conversion of FDG released by AO_MM_FDG to fluorescein by βGal upon 10‐minute irradiation, we observed a slower reaction of βGal encapsulated within AOs (AO_mel_βGal, constant increase over 60 min) compared to its free counterpart (rapid increase within 10 min, Figure S24a, Supporting Information). This difference is attributed to the substrate having to diffuse through pores into the AO cavity to reach the βGal.^{[ 54 ]} As expected, light‐mediated disruption of the AO_MM_FDG is the defining point in the production of fluorescein. When the system was not irradiated, FDG was not released in the surrounding environment, making it inaccessible to βGal for catalysis (Figure S24a, Supporting Information). The release of FDG from AO_MM_FDG is also a restricting factor in the overall enzymatic efficiency of the system. When compared to free given FDG, the increase in fluorescence corresponding to fluorescein was similar to FDG released from AO_MM_FDG, but the overall relative fluorescence increase was lower (Figure S24a,b, Supporting Information). Control AOs without melittin pores (AO_βGal) and without MM (AO_FDG) showed no catalytic activity due to the lack of diffusion and release, respectively (Figure S24b,c, Supporting Information). These results taken together underscore the critical role of the stimulus‐responsive molecular motor in AOs when developing a light‐controlled signaling and communication pathway mimic.

For the development of protocells capable of intracellular signaling, two scenarios were explored: one where βGal was free in the protocell's aqueous cavity and the other, where it was confined in a melittin permeabilized AO (AO_mel_βGal). First, photoreceptive protocells were generated co‐encapsulating free βGal and AO_MM_FDG. After irradiation and subsequent FDG release, a significant increase in fluorescence was measured in the lumen of protocells containing βGal and AO_MM_FDG, thereby creating a single‐compartmentalized intracellular signaling pathway (Figure S25a, Supporting Information). No increase in fluorescence was observed in protocells encapsulating AOs lacking βGal or when not illuminated. We then proceeded to develop the two‐AO protocell by co‐encapsulating AO_mel_βGal and AO_MM_FDG, simulating inter‐organelle signaling. Illumination triggers an intracellular signaling cascade between the two subcompartments in the lumen of the protocell; AO_MM_FDG release FDG, which in turn diffuses into catalytic AO_mel_βGal where it is hydrolyzed to fluorescein, thereby increasing fluorescence. When AO_βGal, which were permeabilized with melittin, were illuminated, the constant increase in fluorescence, together with a significantly higher signal intensity inside the protocell after 1 h, confirmed successful intracellular communication between the two AO populations (Figure 3c; Figure S26, Supporting Information). However, in non‐permeabilized AO_βGal and without illumination or motors, no significant increase in fluorescence was observed (Figure 3c; Figure S25b, Supporting Information).

6. Directional Communication Between Protocell Populations

The task of mimicking a biological signaling network was approached by engineering a compartmentalized setup of two distinct types of protocells (Figure  4 ): a photoreceptive sender protocell containing AO_MM_FDG and a receiver protocell equipped with catalytic AO_mel_βGal organelles. Hierarchical bottom‐up integration of biomolecules and AOs inside two different populations of protocells and their regulation to achieve directionality of the sequential chain of reactions is completely different than the individual behavior of the separate components. The system is functional only when the substrates enclosed inside photo‐responsive organelles (AO_MM_FDG) located inside sender protocells are released upon an external light signal, pass through both sender and receiver protocell boundaries, and finally enter catalytic organelles (AO_mel_βGal) where they reach the encapsulated enzyme. Including MM‐organelles inside sender protocells while the catalytic organelles further downstream in the chain of reactions are located inside receiver protocells enables fine‐tuning in space and time of the molecular transport through various synthetic membranes of the system. The unique sequence of physical and chemical signal propagation across different membranes is intended to mimic authentic signal transduction.

Figure 4.

Intercellular signaling involving organelles spatially confined in separate protocells. a) Schematic overview of light‐triggered signaling cascade from sender to receiver protocell. Irradiation at 430 nm causes of FDG‐encapsulating AOs to rupture, releasing non‐fluorescent FDG. FDG diffuses via DNA nanopores from the sender protocell to the receiver protocell containing AOs encapsulating the enzyme βGal. Inside the receiver protocells, FDG enters AO_mel_βGal via melittin pores, and finally gets hydrolyzed to its fluorescent product fluorescein by the confined enzyme. b) Normalized fluorescence inside receiver protocells encapsulating AO_mel_βGal with and without DNA nanopores after 1 h of co‐incubation with sender protocells encapsulating free FDG (n ≥ 5). c) Normalized fluorescence inside receiver protocell encapsulating AO_mel_βGal with and without DNA nanopores after 1 h of co‐incubation with sender protocells encapsulating FDG in photolabile AOs (AO_MM_FDG) (n ≥ 5). d) Fluorescence micrographs of representative single receiver protocells from (c) showing fluorescein fluorescence (white) without (left) and with (right) DNA nanopores after 1 h co‐incubation with AO_MM_FDG‐encapsulating sender protocells. Scale bar, 10 µm.

Permeabilization was achieved by extracellular administration of DNA nanopores, whose integration into the protocell membrane enabled diffusion of substrates from and to the sender/receiver protocells, thereby establishing an intercellular communication system (Figure 4a). DNA nanopores with an approximate molecular weight cutoff of ≈50 kDa were previously shown to permeabilize PDMS₂₅‐b‐PMOXA₁₀ GUVs and facilitate substrate transfer.^{[ 22 ]} We evaluated their insertion into protocells loaded with non‐compartmentalized βGal by externally adding FDG and monitoring its conversion to fluorescein inside the catalytic protocell (Figure S27, Supporting Information).To selectively label the receiver protocell population, we encapsulated CF633‐dextran (70 kDa) as a high molecular weight fluorescent label too big to exit through the DNA nanopores together with AO_mel_βGal. The dextran concentration was optimized to minimize its interaction with the AOs and with FDG hydrolysis, which occurred particularly at dextran concentrations above 0.1 µm (Figure S28, Supporting Information). When free FDG was encapsulated in the sender protocell, an increase of fluorescence was observed in the receiver protocell only if both, sender and receiver protocells were permeabilized with DNA nanopores (Figure 4b), irrespective of whether the protocells were exposed to light or not. Harnessing sender protocells with photoreceptive properties through encapsulation of AO_MM_FDG rather than free FDG led to a significant increase in fluorescence in receiver protocells containing AO_mel_βGal only upon illumination and provided both, sender and receiver protocells were allowed communication via DNA nanopores (Figure 4c,d). Furthermore, no increase in membrane permeability or transfer of membrane fragments between the AO and membranes was detected in protocells containing melittin‐permeabilized AOs (Figures S20 and S29, Supporting Information). These findings confirm the effective directional organelle‐based chain of reactions mediating cell‐cell communication between distinct protocell populations. By choosing a 6:1 ratio between sender and receiver protocells, signal integration of multiple sender protocells to a single receiver protocell, reminiscent of the photoreceptor signal transduction, was achieved.^{[ 61 ]}

7. Ca²⁺ Modulates Signal Propagation in Communicating Protocells

In native rod and cone photoreceptors, light and calcium play a key role in visual phototransduction.^{[ 52} , ^{62 ]} Analogous to the modulatory effect Ca²⁺ has on photoreceptive cells, we explored the influx and sensing of environmental Ca²⁺ in subcompartmentalized protocells bearing the ionophore ionomycin, which is known to transport Ca²⁺ across membranes.^{[ 63} , ⁶⁴ , ^{65 ]} To this end, the membrane‐impermeable calcium‐sensitive dyes CalciumGreen‐5N (CaGreen) or Rhod‐5N (R5N), whose fluorescence increases upon Ca²⁺ binding (Figures S30 and S31, Supporting Information), were loaded in melittin‐permeabilized AOs (AO_mel_CaGreen and AO_mel_R5N, respectively; Table S1, Figures S32 and S33, Supporting Information). The resulting Ca²⁺ reporting AOs were tested at different Ca²⁺ concentrations for their sensitivity to detect Ca²⁺ both in bulk and when encapsulated inside protocells (Figures S34 and S35, Supporting Information).^{[ 27} , ^{54 ]} Further characterization and optimization of experimental conditions are described in the Supporting Information. To ultimately examine the modulatory effect of Ca²⁺ on the response of receiver protocells containing AO_mel_bGal to incoming signals, the hydrolysis of FDG by βGal, free and encapsulated in AOs (AO_mel_ βGal), was first examined in the presence of 0 to 1 mm CaCl₂ (Figure S36, Supporting Information). As expected, without CaCl₂, the reaction took place rapidly, and a plateau was observed after 2 h. However, 1 mm CaCl₂ decelerated the reaction and a plateau, which was threefold lower, was reached after 3 h. The slopes of the reactions, corresponding to the rate of fluorescein production, revealed a ≈5‐fold decrease in the hydrolysis of FDG by βGal (Figure S36d, Supporting Information), which is in agreement with the reported effect of calcium and heavy metals on βGal activity.^{[ 66} , ⁶⁷ , ^{68 ]} Similarly, in mixtures of sender and receiver protocells, extracellularly added Ca²⁺ also modulated βGal activity in the receiver protocell upon illumination (Figure  5a,b Figure S37, Supporting Information).^{[ 68 ]} Specifically, in the presence of 1 mm CaCl₂, the fluorescein signal in receiver protocells decreased by ≈10% (Figure 5c). No increase in fluorescence was observed in the absence of DNA nanopores, regardless of irradiation or CaCl₂ presence (Figure 5c; Figure S38, Supporting Information). To confirm that the CaCl₂ not only diffuses through the DNA nanopores but can also enter the lumen of the AOs, melittin‐permeabilized AOs containing Rhod‐5N were encapsulated in protocells and the Ca‐dependent fluorescence increase was monitored (Figures S34,S35,S37, Supporting Information). More importantly, the data from communicating protocells showed that the signal propagation between sender and receiver protocell systems could be modulated by CaCl₂.

Figure 5.

Sensitivity modulation of intercellular communication. a) The external addition of CaCl₂ decreases the signal transmission in the receiver protocell, thereby decreasing the gain of the system. b) Calcium has an inhibitory effect on βGal encapsulated in the receiver protocell, thereby modulating the overall photosensitivity of the receiver protocell. c). Normalized fluorescein intensity in receiver protocells encapsulating AO_mel_βGal in the presence or absence of CaCl₂ after 1 h of incubation (n ≥ 5). Significance levels: p > 0.05 (n.s.), p < 0.05 (*), p < 0.005 (**), and p < 0.0005 (***).

8. Conclusion

In summary, bottom‐up engineering resulted in subcompartmentalized protocells able to support an organelle‐based chain of reactions mediating intercellular communication between populations of protocells in response to environmental signals. The controlled encapsulation of AOs through microfluidics allowed a fine tuning of organelle density and inter‐organelle distances. Equipping the membranes of protocells and AOs with ionophores, pores or light‐driven rotary molecular motors facilitated the spatiotemporal control over their response and generated a multi‐stimuli responsive network of protocells by combining functionalities responsive to different external stimuli. Photoresponsive protocells were equipped with molecular motor‐bearing AOs, that underwent structural changes upon light exposure, thereby releasing the AOs' content. Two different types of protocells were developed by encapsulation of substrate‐releasing photoresponsive AOs and enzyme‐based catalytic AOs. Including MM‐organelles inside sender protocells while downstream catalytic organelles are located inside receiver protocells involves an unprecedented degree of complexity and enables refined tuning in space and time of molecular transport through various synthetic membranes of the system. First, by co‐encapsulating photoresponsive and catalytic AOs inside a single protocell, we evaluated the intracellular signal propagation. Then, through the segregation of specific AOs into separate protocells, directional cell–cell communication was achieved between sender protocells responding to light and downstream receiver protocells. Adding Ca²⁺‐responsive AOs to the sender protocells enabled the creation of a dual‐responsive system that simultaneously sensed environmental Ca²⁺ and responded to light input. Through external administration of CaCl₂, the signal transduction has been modulated in the receiver protocells, allowing for a light adaptation of the system similar to that observed in a natural rod or cone cell.

Of particular interest was integrating molecular motors in AOs membranes as it offers the unique advantage of a time‐controlled tuning of the membrane permeability or even membrane disintegration upon the illumination. One could also envisage a spatially controlled modification of membranes or assemblies hosting motors, for example by inserting molecular motors that undergo structural rearrangements (rotation) at different wavelengths, or a regulation (e.g., switching off) of biological macromolecules to which motors are complexed. Depending on the chemical composition of the individual AOs and their protocell hosts, these systems have the intrinsic property of being capable of changing parameters such as kinetic constants or fluxes, which, in turn, affect the resulting functionality. By a straightforward change of the biomolecules inside the organelles and/or the organelle repertoire inside protocells, completely different sequential reactions can be achieved upon the presence of environmental signals, including the controlled production of desired drugs, release of growth factors, or alleviation of oxidative stress. While here we equipped each protocell type with one or two types of AOs as a proof of concept, by co‐encapsulation of various types of AOs inside each protocell, it will be possible to combine in a single system complex intra‐ as well as intercellular chains of reactions to unravel details about natural signaling pathways in a simpler and more accessible manner than in real cells. Importantly, as both the AOs and the protocells are based on amphiphilic copolymers, they can be chemically functionalized to expose at their exterior specific functional groups serving to link them into specific higher order networks, for example in the development of synthetic prototissues with potential in regenerative medicine. In addition, ensuring close proximity of different protocells inside networks will enable to significantly increase the efficiency of the chain of reactions in producing desired compounds and signals in a controlled, stimuli‐responsive manner. Our system offers a basis for further studies of more complex communication networks providing i) protocells competent of collective behavior, and ii) new methods to interface protocells and living cells through engineered communication networks, such as between protocells and bacterial populations.

9. Experimental Section

Reagents

Polyethylene glycol (PEG, M_n = 35000), polyvinyl alcohol (PVA, MW 13000‐23000, 87–89% hydrolyzed), chloroform (99%), anhydrous hexane (95%), fluorescein di‐(β‐D‐galactopyranoside) (FGD), sucrose, sodium chloride, melittin (from honey bee venom, ≥85% by HPLC), CaCl₂, β‐Galactosidase from Escherichia coli (Grade VIII, ≥500 units mg⁻¹ protein), proteinase K (from Tritirachium album), ionomycin calcium salt (Streptomyces conglobatus, ≥98% (HPLC)), Sepharose (4B, 45–165 µm beads diameter), and Whatman Nucleopore Track‐Etched membranes were obtained from Sigma–Aldrich. Calcium Green‐5N (hexapotassium salt, cell impermeant) was purchased from Invitrogen. Rhod‐5N (tripotassium salt) was obtained from AAT Bioquest. Pluronic F‐68 Non‐ionic Surfactant (100X) was obtained from Gibco. BODIPY 630/650 was obtained from Thermo Scientific Inc. Cy5‐labeled melittin was purchased from Biosynth Ltd. ATTO488‐carboxyl was obtained from ATTO‐TEC. Aquapel was obtained from PGW Auto Glass. DNA nanopores were obtained from Tilibit Nanosystems.^{[ 22 ]} All chemicals were used as received unless stated otherwise.

Synthesis of PDMS₂₅‐b‐PMOXA₁₀

Synthesis of the amphiphilic diblock copolymer poly(dimethylsiloxane)₂₅‐block‐poly(2‐methyl‐2‐oxazoline)₁₀ (PDMS₂₅‐b‐PMOXA₁₀) was described elsewhere.^{[ 37 ]}

Synthesis of Molecular Motor

Detailed procedures for the building block synthesis of the molecular motor are described in references^{[ 69} , ⁷⁰ , ⁷¹ , ^{72 ]} and detailed description of the formation of the motors in reference.^{[ 31 ]} The analytical data is in accordance with the literature (see Supporting Information for more details). All reactions were carried out in flame‐dried Schlenk tubes or oven‐dried crimp top vials under a nitrogen atmosphere using standard Schlenk techniques. Solutions and reagents were added with nitrogen‐flushed disposable syringes/needles. Analytical thin‐layer chromatography was performed on silica gel 60 G/UV265 aluminum sheets from Merck, Germany (0.25 mm). Flash column chromatography was performed on silica gel Davisil LC60A (Merck type 9385, 230 to 400 mesh) or a Reveleris X2 Flash Chromatography system from Büchi (Switzerland) Medium‐Pressure Liquid Chromatography (MPLC).

Microfluidic Device Fabrication and Coating

Microfluidic six‐way junction chips were used as published previously.^{[ 45 ]} Silicon–glass microfluidic chips were etched into silicon wafers at the Binnig and Rohrer Nanotechnology Center at IBM Research Europe – Zurich using deep‐reactive ion etching. The etched microfluidic chips were covered with borofloat 33 (BF33) glass covers comprising pre‐milled trough‐holes as outlets by anodically bonding them. The chip coating and regeneration was carried out according to a published protocol.^{[ 29 ]}

Microfluidic Setup

All fluids were filtered using appropriate hydrophilic or hydrophobic syringe filters (pore size 0.2 µm) prior to passing them through the microfluidic chips. Hamilton gas‐tight syringes (0.5, 1, and 5 mL) with PTFE‐Luer‐lock mechanisms and a Cetoni neMESYS low pressure syringe pump system with 3 modules was used to inject the fluids into the microfluidic chips. Fluorinated ethylene propylene (FEP) tubing (BGB Analytik) with an inner diameter of 0.25 mm (1/16″) were used to connect the syringes to the microfluidic chips.

Double Emulsion Fabrication and Dewetting

PBS supplemented with 200 mm sucrose was used as an inner aqueous phase. The PDMS₂₅‐b‐PMOXA₁₀ was dissolved at 4 mg mL⁻¹ in a 3:2 v/v mixture of hexane and chloroform. For the outer aqueous phase, PBS was supplemented with 5% PEG_35000, 0.1% Pluronic F‐68 and 100 mm NaCl to decrease the osmolarity difference to the IA. A freezing point osmometer (Gonotec Osmomat) was used to measure the osmolarity. Flow rates of 1, 3, and 5 µL min⁻¹ were used to flow the polymeric organic, the inner and outer aqueous phases into the microfluidic chip. After double emulsion formation in the chip, the formed emulsions were collected for 10 min in a 1.5 mL Eppendorf tube containing 300 µL outer aqueous phase. Evaporation of the organic phase yielded ≈10⁶ GUVs at a concentration of 10⁶ mL⁻¹.^{[ 29} , ^{45 ]}

Preparation of AOs

AOs were prepared using the film rehydration method. Briefly, a thin film of PDMS₂₅‐b‐PMOXA₁₀ (5 or 10 mg mL⁻¹ polymer in EtOH) was formed by rotary evaporation. For AOs containing the molecular motor (AO_MM_A488, AO_MM_FDG), the motor was dissolved in EtOH and mixed with the polymer solution (25 mol% MM relative to polymer). For ATTO488‐containing AOs, the film was rehydrated (overnight, stirring, RT) with PBS containing ATTO488‐carboxyl (100 µm). For Cy5‐melittin AOs, the film was rehydrated with Cy5‐melittin (10 or 50 µm) in PBS. For FDG‐containing AOs, the rehydration was performed with FDG (50 µm) in PBS. For AO_CaGreen, the rehydration was performed with CalciumGreen™ (CaGreen) in MilliQ water (100 µM) and melittin (10 µm), for the case of AO_mel_CaGreen. For generating AO_mel_R5N, MilliQ, water containing the fluorescent dye Rhod‐5N (100µm) and melittin (10 µm) was used for film rehydration. For AOs containing β‐Galactosidase (βGal), the film was rehydrated with a solution of βGal in PBS (100 U mL⁻¹) and melittin (50 µm), for AO_mel_βGal. The enzyme‐containing AOs were incubated with proteinase K (0.05 mg mL⁻¹) for 1 h at 37 °C. AOs were extruded through a 200 nm Whatman Nuclepore polycarbonate membrane, followed by purification by size exclusion chromatography (SEC, Sepharose). All samples were kept at 4 °C until further use.

Molecular Motor Activation

AOs containing molecular motors in their membranes were irradiated using a self‐made illumination device equipped with 6 Lite‐On LEDs (Mouser Electronics, 3.7 V, 50 Hz, 0.5 A, 430 nm) with distances of 3.5 cm between them.^{[ 31 ]} Device dimension were 13 × 9 × 3.5 cm. The light intensity was estimated to be ≈31.8 W m⁻² in the sample.

Preparation of AOs and AO‐Encapsulating Protocells

For preparation of AO‐loaded protocells, AOs were directly added to the inner aqueous phase prior to double emulsion production. An overview over the employed AO concentrations can be found in Table S2 (Supporting Information). Protocells were produced under the same flow conditions as non‐AO‐encapsulating protocells.

Protocell Permeabilization Using DNA Nanopores

DNA nanopores were reported elsewhere and were used as purchased.^{[ 22 ]} DNA nanopore concentration was optimized using free β‐galactosidase in protocells and free FDG added from the outside (Figure S27, Supporting Information). The pores were typically added to the protocell suspension 1 h prior to imaging at a final concentration of 1 nm in outer aqueous phase.

Protocell‐Protocell Communication

For protocell communication assays, the sender protocells were diluted in OA in a 1:3 ratio and subsequently illuminated for 10 min at 430 nm in an Ibidi µ‐slide 18 well #1.5 glass bottom coverslip (Ibidi, USA). After illumination, receiver protocells were added to a final dilution of 1:20 (final ratio sender:receiver protocell 1:6) and DNA nanopores were added to a final concentration of 1 nm. Protocell populations were imaged after 1 h incubation at room temperature. For imaging, two frames were integrated to achieve an increased signal to noise ratio.

Protocell Imaging

Protocells were typically diluted 1:10 in OA for imaging by confocal laser scanning microscopy (CLSM) and placed in an Ibidi µ‐slide GUV membranes were, unless otherwise stated, stained with 2.5 µm BODIPY 630/650 hydrophobic membrane selective dye. AOs were typically imaged undiluted on glass cover slide (#1.5). A Zeiss 880 confocal laser scanning microscope (Zeiss, Germany) equipped with a Plan‐Apochromat 20×/0.8 M27 or water‐immersion C‐Apochromat 40×/1.2 objective was employed for image acquisition. A 488 nm argon laser, a 561 nm DPSS 5561‐10 laser, and a 633 nm HeNe laser were used for visualizing fluorophores. Images were typically recorded with an image size of 1024 × 1024 or 2048 × 2048 pixels with the pinhole set to 1 Airy unit and the bit depth set to 16 bit. Images were analyzed using Fiji image analysis software and python scripts to measure GUV fluorescence or size. For AO‐loaded protocells, airy scan z‐stacks were recorded. 3D image reconstruction was done using the Imaris software (Bitplane) software. For stability studies, 11 µL (undiluted) GUV suspension was added to a urinalysis slide (Fast Read 102 counting chambers) and imaged on an Olympus EP50 microscope. 20 µm calcein was added to the IA of the vesicles during production for automated image analysis.

Statistical Analysis

For comparative analysis, two‐way Analysis of Variance (ANOVA) were used, followed by post‐hoc Tukey's honestly significant difference (HSD) testing. Data was tested for their normal distribution prior to statistical analysis using the Shapiro–Wilk test. p < 0.05 was considered statistically significant. The significance level of the calculated p values was indicated using asterisks: p > 0.05 (n.s.), p < 0.05 (*), p < 0.005 (**), and p < 0.0005 (***). Data was normalized to the control dataset and unless otherwise stated, the mean ± standard deviation is presented. In boxplots, dots indicate single samples/GUVs, the center line depicts the median and boxes span the interquartile range (25–75%). Outliers are indicated in the dot plots.

Code Availability

The custom scripts in this study are available at https://github.com/lukasheuberger/vesicleimaging/.

Conflict of Interest

The authors declare no conflict of interest.

Author Contributions

L.H. and M.K. contributed equally to this work. L.H., M.K., and C.G.P. conceptualized the study. L.H., M.K., A.G., and D.M. performed the experiments and conducted investigations. L.H., M.K., and C‐A.S. wrote the manuscript. D.M.L. and E.L. fabricated the microfluidic devices. L.H., M.K., A.G., D.D., and B.L.F. provided resources. C.G.P. and B.L.F. supervised the work, administered the project and acquired funding. All authors contributed to the discussion of the results and reviewed the manuscript.

Supporting information

Supporting Information

Supplemental Movie 1

Supplemental Movie 2

Heuberger L., Korpidou M., Guinart A., Doellerer D., López D. M., Schoenenberger C.‐A., Milinkovic D., Lörtscher E., Feringa B. L., Palivan C. G., Photoreceptor‐Like Signal Transduction Between Polymer‐Based Protocells. Adv. Mater. 2024, 37, 2413981. 10.1002/adma.202413981

Data Availability Statement

The data that support the findings of this study are openly available in Zenodo at https://doi.org/10.5281/zenodo.10778462, reference number 10778462.