Self-Regulated and Bidirectional Communication in Synthetic Cell Communities

Abstract

Cell-to-cell communication is not limited to a sender releasing a signaling molecule and a receiver perceiving it but is often self-regulated and bidirectional. Yet, in communities of synthetic cells, such features that render communication efficient and adaptive are missing. Here, we report the design and implementation of adaptive two-way signaling with lipid-vesicle-based synthetic cells. The first layer of self-regulation derives from coupling the temporal dynamics of the signal, H₂O₂, production in the sender to adhesions between sender and receiver cells. This way the receiver stays within the signaling range for the duration sender produces the signal and detaches once the signal fades. Specifically, H₂O₂ acts as both a forward signal and a regulator of the adhesions by activating photoswitchable proteins at the surface for the duration of the chemiluminescence. The second layer of self-regulation arises when the adhesions render the receiver permeable and trigger the release of a backward signal, resulting in bidirectional exchange. These design rules provide a concept for engineering multicellular systems with adaptive communication.

Keywords: adaptive; bidirectional communication; chemiluminescence; self-regulation; synthetic cells.

Figure 1

Design of self-regulated and adaptive two-way communication between synthetic cells. Step 1: glucose and luminol enter the sender GUV (green) through α-hemolysin (α-HL) pores. Here, glucose oxidase (GOx) converts glucose into intracellular H₂O₂ (Reaction 1), which serves a dual purpose: photoactivation of the adhesions between sender and receiver GUVs in Step 2 and as a chemical signal in Step 3. Through the HRP-catalyzed chemiluminescence reaction between H₂O₂ and luminol, light (Reaction 2) is produced inside the sender GUV, which activates the photoswitchable protein iLID* at the surface (Reaction 3). Step 2: the sender GUV adheres to the receiver GUV (red), which is functionalized with Nano at its surface due to the binding between iLID* and Nano (Reaction 4), and brings the GUVs within signaling range. Step 3: the proximity of the GUVs and the increased permeability of the receiver GUV upon adhering to the sender GUV with α-HL pores together enable efficient two-way signaling between GUVs. Forward communication from sender to receiver through H₂O₂ results in an increased response from the sensor roGFP-Orp1, and backward communication from receiver to sender through Ca²⁺ results in an increased response from the sensor CalGreen (Reaction 5). Step 4: sender and receiver GUVs separate after two-way communication, as H₂O₂ production and chemiluminescence decreases and the interaction between iLID and Nano reverses in the dark (Reaction 6).

Figure 2

Photoactivation of iLID by chemiluminescence. (a) Schematic representation of heterotypic GUV-GUV adhesions triggered by chemiluminescence. Luminol and H₂O₂, membrane-permeable substrates, generate chemiluminescence in sender GUVs (green) catalyzed by encapsulated HRP. The generated light activates iLID, which is immobilized on the GUV membrane. The senders with photoactivated iLID* adhere to the receiver GUVs (red) that are functionalized with Nano and bind. (b) The absorption spectrum (black line) of iLID (20 μM) and the chemiluminescence emission spectrum (orange line) of the HRP (2 μg/mL)-catalyzed chemiluminescence reaction between luminol (200 μM) and H₂O₂ (500 μM) spectrally overlap. (c) The chemiluminescence over time in the absence and presence of iLID. Reaction conditions are as in (b). (d) Competitive fluorescence polarization assay, where the binding of the SsrA-TAMRA peptide (200 nM) to Nano (1 μM) increased fluorescence polarization. SsrA-TAMRA was competitively displaced when iLID (5 μM) was photoactivated through the chemiluminescence reaction (same conditions as in (b)), resulting in a decrease in fluorescence polarization. Later, Nano dissociated from iLID and bound to SsrA-TAMRA again as the photon output of the chemiluminescence reaction decreased, which was visible as an increase in fluorescence polarization. Control experiments in which one of the reagents for the chemiluminescent reaction was omitted did not show significant changes in fluorescence polarization over time. (e) Chemiluminescence inside GUVs loaded with HRP (2 μg/mL) (orange line) compared with the exterior buffer (gray line) as membrane-permeable substrates luminol (200 μM) and H₂O₂ (500 μM) were added externally. (f) Fluorescence confocal microscopy images of sender GUVs (membrane labeled with DiL, shown in green; loaded with HRP; surface functionalized with iLID) and receiver GUVs (membrane labeled with DiD, shown in red; surface functionalized with Nano) with chemiluminescence (CL) (external addition of luminol (200 μM) and H₂O₂ (500 μM)) in the dark and under blue illumination without luminol or H₂O₂ after 30 min. The chemiluminescence sample was repeated with HRP loaded in receiver GUVs as the control of activation by intercellular chemiluminescence. The scale bar is 30 μm. (g) Quantification of the average cluster size, where connected GUVs were analyzed as single objects. Heteroclustering was assessed by merging the DiL and DiD channels, and homoclustering was assessed by analyzing two channels separately. Experiments were performed in three independent technical replicates, and a total area of 0.5 mm² was analyzed for each sample. The results are reported as the average of the mean ± standard error; statistical significance was evaluated by one-way ANOVA, ns >0.05, 0.01 < *p < 0.05, ****p < 0.0001.

Figure 3

Coupling intracellular signal (H₂O₂) generation to spatial proximity of sender and receiver GUVs through chemiluminescence. (a) Schematic representation of the sender GUV (green) that generates H₂O₂ from glucose through the enzymatic activity of GOx. The coupled HRP-catalyzed chemiluminescence reaction photoactivates iLID on the surface of the sender GUV and results in adhesions to the Nano-functionalized receiver GUV (red). As the substrates are consumed, H₂O₂ production and consequently luminescence in the sender GUV decreases, which leads to the deactivation of iLID and the detachment of sender and receiver GUVs. Consequently, the spatial proximity of sender and receiver become a function of intracellular signal generation. (b) Confocal fluorescence microscopy images of GUVs (membrane labeled with DiL, shown in green) capable of generating H₂O₂ (100 mU/ml GOx, 15 ng/μL α-HL pores) loaded with the ratiometric H₂O₂ sensor roGFP-Orp1 (ratiometric signal shown in cyan). The ratiometric signal from roGFP-Orp1 increased over time upon addition of external glucose (20 mM). The scale bar is 20 μm. (c) Quantification of roGFP2-Orp1 response over time. The mean ratiometric response and the standard error of 5 GUVs were reported. (d) Chemiluminescence kinetics in GUVs (15 ng/μL α-HL pores) loaded with GOx (100 or 500 mU/mL) and HRP (2 μg/mL) upon addition of luminol (200 μM) and glucose (50 mM). (e) Fluorescence confocal microscopy images of sender GUVs (membrane labeled with DiL, shown in green; loaded with GOx (100 mU/ml or 500 mU/ml), HRP (2 μg/mL), and α-HL (15 ng/mL); surface functionalized with iLID) and receiver GUVs (membrane labeled with DiD, shown in red; surface functionalized with Nano) reversibly clustering with chemiluminescence and in the dark as a control over time. Reagents were as in (d). The scale bar is 30 μm. (f) Quantification of the average object size, where connected GUVs were analyzed as single objects. Experiments were done in three independent technical replicates, and a total area of 0.5 mm² was analyzed for each sample.

Figure 4

Chemiluminescence-dependent proximity regulates forward H₂O₂ communication within GUV communities. (a) Schematic representation of H₂O₂ produced by GOx from glucose inside the sender GUV (green), serving a dual purpose. H₂O₂ is used in the HRP-catalyzed chemiluminescence reaction to photoactivate iLID, and as a result, the sender GUV adheres to the receiver GUV (red). These adhesions bring the receiver GUV into signaling range, such that H₂O₂ can also serve as a signaling molecule, visible through the ratiometric H₂O₂ sensor roGFP2-Orp1. (b) Confocal fluorescence microscopy images of sender GUVs (membrane labeled with DiL, shown in green; loaded with GOx, HRP, and α-HL; surface functionalized with iLID) and receiver GUVs (membrane labeled with DiD, shown in red; loaded with 20 μM roGFP2-Orp1; surface functionalized with Nano) 10 min after initiating the chemiluminescence reaction through the addition of glucose and luminol. In controls, functionalization with the proteins iLID and Nano or the loading of GOx for H₂O₂ generation was omitted. All reagents were as in Figure 3d. The scale bar is 30 μm. (c) Quantification of the ratiometric signal from the H₂O₂ sensor roGFP2-Orp1 in receiver GUVs proximal and distal to a sender GUV in (b). Experiments were done in three independent technical replicates, and 30 GUVs were analyzed for each condition. The results are reported as the average of the mean and as standard error; statistical significance was evaluated by one-way ANOVA, ****p < 0.0001.

Figure 5

Self-regulated H₂O₂ forward communication leads to adaptive backward communication through Ca²⁺. (a) Confocal fluorescence microscopy images of sender GUVs (membrane labeled with DiL, shown in green; loaded with α-HL; surface functionalized with iLID) and receiver GUVs (membrane labeled with DiO, shown in cyan; surface functionalized with Nano) loaded with sulfo-Cy5 (shown in red). Proximal receiver GUVs become permeable due to α-HL pores in sender GUVs and leak the loaded sulfo-Cy5, but distal receiver GUVs retain the loaded sulfo-Cy5. Contacts between the GUVs were induced under blue light for 30 min. The scale bar is 30 μm. (b) Quantification of sulfo-Cy5 fluorescence in (a). Thirty GUVs were analyzed for each population. The results are reported as the average of the mean and as standard error; statistical significance was evaluated by one-way ANOVA, ****p < 0.0001. (c) Schematic representation of adaptive two-way communication between sender and receiver GUVs. H₂O₂ produced by GOx from glucose inside the sender GUV (green) serves both to bring the sender and receiver GUVs in close proximity and as a forward chemical signal from the sender to the receiver GUV, visible through the H₂O₂ indicator roGFP2-Orp1 (cyan). The receiver GUV becomes more permeable due to α-HL in the sender GUV upon contact. Consequently, the backward signal from the receiver to the sender GUV, Ca²⁺, is released and received in the sender GUV, visible as an increase in the Calcium Green C24 signal in the membrane (purple). (d) Confocal fluorescence microscopy images for the setup described in (c). Sender GUVs (membrane labeled with DiL, shown in green, and Calcium Green C24, shown in purple; loaded with GOx, HRP, and α-HL; surface functionalized with iLID) and receiver GUVs (membrane labeled with DiD, shown in red; loaded with 20 μM roGFP2-Orp1, shown in cyan; surface functionalized with Nano) after 30 min of adding the enzyme substrates. The scale bars are 20 μm. (e) Quantification of H₂O₂ sensor in receiver GUVs and Ca²⁺ sensor in sender GUVs distal and proximal to each other. Experiments were done in three independent technical replicates, and 30 GUVs were analyzed for each population. The results are reported as the average of the mean and as standard error; statistical significance was evaluated by one-way ANOVA, ****p < 0.0001.