Strategies and applications of synthetic cell communication

Abstract

Cell signalling and communication are fundamental to living cellular communities. For the past two decades, there has been continuous development of bottom-up engineered synthetic cells, which have become more and more similar to their natural counterparts. However, we are only scratching the surface with the development of synthetic cellular communities and their integration into natural tissues. Here, we review different intercellular communication mechanisms engineered for synthetic cells and classify them based on their resemblance to natural cell signalling mechanisms: autocrine, paracrine, and juxtacrine. In particular, we highlight recent advances in molecular tools for intercellular communication designs and discuss potential applications of engineering synthetic cellular communities and synthetic cell-natural cell communication. With further advances in this area, synthetic cellular communities will be powerful tools for understanding and manipulating cellular functions, thus unlocking potential applications in biosensing, cellular reprogramming, and sustainability.

Introduction

Imagine a world in which engineered cells, plants, tissues, and non-living yet biologically active entities are integrated into our everyday lives and play critical roles in solving grand challenges in health, biotechnology, and sustainability. In the eyes of a bioengineer, this will include the exploitation of both top-down synthetic cells, made by manipulation and engineering of natural cells, and bottom-up synthetic cells, non-living entities constructed by mixing molecules of life, which hold promise as solutions to various environmental and biomedical challenges.

Bottom-up synthetic cells, as we define them in this review and refer to them as synthetic cells hereafter, are compartmentalized biologically smart reactors engineered to serve specific purposes. These micron-sized, spatially bound compartments encompass various biochemical molecules that enable them to sense external stimuli, process the signals via chemical reactions, and lastly, respond via designed biochemical processes to realize communication with other synthetic cells or with living cells (Figure 1).^1,2

Figure 1∣. Synthetic cells and their current tools for intercellular communication.

a. Major development of their signal sensing, processing, and responding capabilities has allowed the development of communication between different types of synthetic cells as well as between synthetic cells and living cells. b. A vast array of molecular tools have been developed to enable this communication.

Since their first demonstration, many sophisticated biologically inspired processes have been reconstructed in synthetic cells, thus making them increasingly similar to their natural counterparts. Modern synthetic cells can be designed to display many cell-like behaviors ranging from shape change,³ division,⁴ and motility^5,6 to exocytosis,⁷ DNA replication,⁸ and photosynthetic energy regeneration.⁹ While some of these capabilities were bestowed upon synthetic cells by simply isolating the responsible cellular machinery (such as actin and actin-binding proteins in the cytoskeleton¹⁰) and encapsulating them in vesicles, biomimicry of many biological processes was made possible by utilizing either repurposed or fully synthetic biomolecules. Transmembrane signalling via synthetic receptors,¹¹ DNA-based scaffolds,¹² and molecular transport across the membrane through synthetic molecules¹³ or DNA pores^14,15 are examples of reconstructing a biological event in synthetic cells via non-natural methods.

Despite these advances, in most studies, synthetic cells have remained “single cells”. Communication and cooperation between cells, even among primitive prokaryotes, is critical in the evolution of multicellular systems as well as the growth and survival of communities of single cells. Through signalling and communication, cells perceive their population size and the presence of competing species,¹⁶ display collective emergent functions,^17,18 and form tissues composed of many different cell types.¹⁹ While prokaryotes often communicate through diffusible chemical molecules, metazoans have amassed a rich repertoire of communication mechanisms ranging from contact-dependent signalling between juxtaposed cells to endocrine signalling via signals that travel a long distance using a circulatory system.

The impacts of synthetic cells on biomedical and environmental applications, therefore, can be unlocked if synthetic cells are equipped with signalling and communication capabilities that enable the formation of smart communities and tissues of synthetic cells collaborating with each other towards a designed goal.²⁰ Endowing synthetic cells with communication mechanisms also allows synthetic cell-natural cell interfaces, thus opening up new avenues in applications in drug delivery and biomaterials.^21,22

Here, we describe the state of the art in engineering signalling and communication among synthetic cells. We briefly introduce the common chassis of synthetic cells, then review the advances in implementing synthetic cell communication systems and classify them based on the three major bio-inspired categories of autocrine, paracrine, and juxtacrine signalling. We note that natural cells also rely on other signalling mechanisms such as endocrine, intracrine,²³ and ephaptic coupling²⁴ (i.e., an electric field effect found in the nervous system and the heart). However, since the state of the art is focused mainly on the three described categories, we will limit the scope of this review to those mechanisms.

Chassis of synthetic cells

At their most basic, living cells consist of an aqueous solution surrounded by a lipid bilayer membrane. This description is that of a vesicle, which in itself, cannot be described as a synthetic cell. A synthetic cell should have at least one cellular function (or a mimic of). This opens up a wide array of synthetic cell types that can carry out a multitude of functions, such as gene expression, membrane signalling, or communication (Figure 1a).

Giant unilamellar vesicles (GUVs), defined as having a diameter of > 1 μm and therefore ‘cell-sized’, are the most widely used synthetic cell model, as their membranes can be composed of the same lipids as living cells and allow the incorporation of natural membrane proteins. GUVs can be prepared by many different routes, the two predominant methods include emulsion phase transfer and microfluidics.²⁵ Both of these methods allow for the incorporation of proteins within the interior, and asymmetric lipids in each leaflet of the membrane. Aqueous compartments bounded by a lipid monolayer have also been made within an oil solution.²⁶ While these droplets are in a much less ‘cell-like’ environment, their advantage is that they can be brought together to make ‘droplet interface bilayers’ (DIBs), allowing direct connection of two to thousands of droplets in predefined patterns.

Lipid bilayer membranes are susceptible to rupture and have limited permeability, often requiring pore proteins to transport molecules across the membrane. There are alternatives to GUVs that overcome these limitations. Polymersomes are compartments where the bilayer is made of amphiphilic block-co-polymers instead of lipids.²⁷ Their membranes are thicker and more rigid, giving them greater stability, however, this means they are less permeable than GUVs. To create permeable membranes, proteinosomes have been generated using protein-polymer conjugates, which allow the movement of small molecules, nucleic acids, and proteins.²⁸

Moving beyond membrane-bound aqueous interiors, alternative synthetic cells can be generated from membrane-less compartments, such as coacervates and hydrogels. Coacervates are formed from liquid-liquid phase separation of many different types of molecules, including polyelectrolytes, biomolecules such as DNA or RNA, and surfactants.^29,30 They have been shown to selectively uptake biomolecules, including nucleic acids, but are limited in the conditions they can tolerate. Hydrogels are water-swollen molecular networks composed of synthetic or natural polymers.³¹ As these are chemical networks, almost any group or molecule can be included in hydrogels to enable new functionality. Regardless of the type of synthetic cell chassis, developing robust communication schemes between synthetic cells is of great interest.

Molecular tools for synthetic cell communication

To realize distinct communication mechanisms between synthetic cells, there has been a drive to develop molecular tools to allow the detection, processing, and response of communication signals (Figure 1b). Due to the simplicity of fluorescence as a readout, often the response is the production and detection of a fluorescent molecule, which reports on the performance of the system. However, applicable communication mechanisms are now coming to the fore, thanks to new molecular signalling tools.

In synthetic cells, communication often results in the activation of enzymatic pathways within the compartment or at the membrane. The simplest detection systems rely on a single enzyme, for instance, glucose oxidase, which can detect the presence of glucose and respond by generating hydrogen peroxide that changes the pH of the lumen. This pH change is often detected by the generation of a fluorescent molecule in conjunction with horse radish peroxidase.^32-35 Enzymatic pathways can incorporate allosteric amplification to increase the sensitivity for incoming signals, although their practical demonstration has remained limited.³⁶

Alternatively, a common, but more cell-like system, relies on the incorporation of cell-free protein synthesis (CFPS) to carry out transcription and translation of a DNA template to a protein. Often the DNA template just encodes for a fluorescent protein,³⁷ but there are many examples of more applicable outputs, including toxins to kill cancer cells,³⁸ quorum synthetases to produce bacterial signalling molecules,³⁹ and mammalian differentiation factors.⁴⁰ Pore proteins are also frequently used to permeabilise a lipid membrane to release or allow entry of molecules^41-43 (See ref. 44 for an overview of different pores for membrane permeabilization). In a mimic of enzymatic pathways, compartmentalized polymerase-endonuclease-nickase DNA reactions created reaction networks with accelerated reaction kinetics.⁴⁵ The use of DNA nanotechnology can be adopted as a communication tool within synthetic cells, for instance using toehold-mediated strand displacement⁴⁶ or DNA nanopores.^15,47,48 The inclusion of DNA strands on the outer membrane of synthetic cells can also be used to induce the fusion of separate compartments.^49,50

There is a vast array of molecular signals that can activate molecular processes within synthetic cells. This often relies on biological molecules that can be recognised by the pathways included, such as ATP, sugars, homoserine lactones, or similar synthetic versions, such as theophylline and IPTG. Using more porous membranes or membraneless compartments, it is also possible to detect short DNA strands or proteins as the signal molecule.^29,46 Beyond small molecules, there is also a drive to develop remote control mechanisms to activate synthetic cells using physical stimuli. Light-^51,52 and temperature-sensitive lipids⁵² can be used to induce permeability or rupture of synthetic cells or organelles within. The inclusion of light-activated DNA^53,54 or proteins⁵⁵ and temperature-sensitive RNA⁵⁶ within the compartments can be used to activate the internal processes. Physical remote stimuli offer a major advantage over small molecules in that they can cross the membrane of any compartment and can offer spatiotemporal resolution in activation.

Mechanisms for biomimicry of intercellular signalling

Common approaches to engineering intercellular signalling and communication among synthetic cells have been heavily influenced by how nature implements intercellular signalling. Different methods for mimicking autocrine, paracrine, and juxtacrine signalling have been developed. Some methods are fully synthetic while the rest are natural biochemical components reprogrammed for intercellular communication. Among the three categories of intercellular signalling, examples of paracrine and autocrine signalling are ubiquitous in literature, mainly because engineering juxtacrine signalling, despite its unique advantages, requires specialized molecular tools such as transmembrane proteins. In this section, we will review the state-of-the-art engineered mechanisms of synthetic cell communication categorized based on their resemblance to natural cell-cell signalling schemes. A summary of the studies on intercellular communication for synthetic cells according to the three abovementioned categories is presented in Table 1.

1. Autocrine
Synthetic Cell Type	Signal	Signal detection mechanism	Receiver Cell Response	Novelty	Reference
Gel-based compartments in water-in-oil emulsion droplet	mRNA	Sequence-specific hybridization	mVenus expression	Implementation of gel-based compartments that enable spatial control of reactions within a synthetic cell system, emulating eukaryotic cellular complexity	Aufinger et al.58
Phospholipid vesicles containing LUVs and internalized large phospholipid vesicles as organelles	Glucose as the external signal and hydrogen peroxide as the autocrine signal	Enzymatic reaction: HRP catalyzes the conversion of the Amplex Red substrate into the fluorescent molecule resorufin	Red fluorescence production through conversion of Amplex Red to resorufin	Demonstration of chemical autocrine signalling between two distinct organelles within a lipid-based synthetic cell	Li et al.35
DNA-based coacervate compartments in water-in-oil emulsion droplet	UV/Vis light as the external signal and ssDNA as the autocrine signal	ssDNA hybridization with complementary sequence residing in each compartment	Colocalization of signals’ green fluorescence with the compartments’ blue/red fluorescence marker	Using LLPS of DNA coacervates as organelles in synthetic cells and controlling autocrine signalling by external UV/visible light	Zhao et al.59
Phospholipid vesicles containing LUV-based organelles	Stachyose as the external signal and glucose as the autocrine signal	Two enzymatic reactions in separate organelles converting stachyose to glucose and glucose to gluconolactone	Red fluorescence production through conversion of Amplex Red to resorufin	Using a microfluidic platform for the hierarchical construction of phospholipid-based synthetic cells with a controlled number of encapsulated organelles	Shetty et al.60
Proteinosome compartments inside gel-based coacervate vesicles	Glucose and urea as external signals and hydrogen peroxide as the autocrine signal	Converting o-phenylenediamine by HRP into the fluorescent molecule 2,3-diamino phenazine (2,3-DAP)	Green fluorescence response	Construction of symbiotic nested proteinosome organelles in coacervate vesicles with their morphology being regulated by external signals and also autocrine communication between organelles	Yin et al.61
2. Paracrine
Sender/Receiver Cell Type	Signal	Signal detection mechanism	Receiver Cell Response	Novelty	Reference
Phospholipid vesicles → Phospholipid vesicles	Arabinose/theophylline as the external signal and IPTG / doxycycline as the paracrine signal	Activation of gene expression	Bioluminescence	Demonstration of signalling through diffusible inducers between synthetic cells encoded with gene expression cassettes for response generation	Adamala et al.42
	Adenosine 5'-triphosphate (ATP) as the external signal and adenosine 5'- monophosphate (AMP) as the paracrine signal	Allosteric activation of glycogen phosphorylase b (GPb)	Production of NADH from glycogen	Implementation of a signal amplification design in receiver cells where GPb becomes catalytically active through allosteric binding of the signalling molecule, AMP	Buddingh’ et al.36
Porous microcapsules → Porous microcapsules	T3 RNA polymerase	Gene expression and tight binding of TetR with TetO	Accumulation of TetR-GFP in the nucleus of the receiver cell	Construction of a clay-DNA hydrogel as the synthetic cell nucleus and use of a porous polymer as the membrane of synthetic cells allowing signalling using a large molecule	Niederholtmeyer et al.67
Phospholipid vesicles → Proteinosomes	3OC6-HSL as the external signal and glucose as the paracrine signal	Oxidation of glucose by GOx	Red fluorescence production through conversion of Amplex Red to resorufin	Demonstration of communication between synthetic cells of different types. Paracrine signal release as a result of gene expression in sender cells	Tang et al.68
Phospholipid vesicles → Eukaryotic cell	3OC6-HSL as the external signal and BDNF as the paracrine signal	Activation of tropomyosin receptor kinase B (TrkB)–BDNF signalling	Expression of differentiation factors or GFP	Optimization of synthetic cells for coculture and communication with mammalian cells. Reconstitution of the large bacterial pore, perfringolysin O, to enable protein-based signalling	Toparlak et al.40
Phospholipid vesicles → Bacteria	Light as the external signal and N-isovaleryl-L-homoserine lactone (IV-HSL) as the paracrine signal	Transcriptional regulation via binding of BjaR with IV-HSL	GFP expression	Encapsulating the enzymatic reaction for production of membrane-permeable IV-HSL in sender cells and development of BjaR gene circuit in receiver cells with high sensitivity for IV-HSL	Smith et al.53
	Theophylline as the external signal and IPTG as the paracrine signal	Transcriptional regulation through IPTG-induced expression of T7 RNA polymerase	GFP expression	First demonstration of synthetic cell – living cell communication. Synthetic cells translate an undetectable external signal to a signal that E. coli responds to	Lentini et al.87
Bacteria ↔ Phospholipid vesicle	Quorum sensing molecules 3OC6-HSL, C8-HSL, and 3OC12-HSL	Transcriptional regulation by sensing the autoinducer molecules	Bioluminescence and GFP expression	Engineering synthetic cells to act as mediators between two bacterial species by sensing paracrine signals from one and sending different paracrine signals to the other	Lentini et al.39
Proteinosomes → Proteinosomes	A trigger ssDNA as the external input and ssDNA as paracrine signal	Toehold-mediated strand displacement	Dequenching of a fluorophore within the receiver proteinosomes	Using toehold-mediated DNA strand displacement to encode computation and logic inside receiver cells.	Joesaar et al.46
	Light as the trigger for signalling and ssDNA as the paracrine signal	Toehold-mediated strand displacement	Dequenching of a fluorophore within the receiver proteinosomes	Inclusion of a photocleavable spacer in the dsDNA sequence inside sender cells enables spatially precise signal release	Yang et al.70
	ssDNA as the external signal and RNA as the paracrine signal	Toehold- or dCas9-mediated strand displacement and Cas12a DNA editing	Dequenching of a fluorophore within the receiver proteinosomes	Demonstration of using RNA as a paracrine signal and using a CRIPSR/Cas system to process the signal within receiver cells	Yang et al.71
3. Juxtacrine
Sender/Receiver Cell Type	Signal	Signal detection mechanism	Receiver Cell Response	Novelty	Reference
Phospholipid vesicles → Phospholipid vesicles	Coelenterazine as the external signal and light generated by luciferase inside sender cells	Photoactivation of iLID and adhesion of sender cell to receiver cell	Lysis induced by activation of PLA2	Demonstration of light-mediated adhesion between proximal synthetic cells that enabled chemical communication through Ca2+ between sender and receiver cells	Chakraborty et al.80
	Furimazine as the external signal and light generated from NanoBit at vesicle-vesicle interfaces as the juxtacrine signal	Photoactivation of membrane-bound iLID inside receiver cells	Recruitment of fluorescent SspB to the contact interface of sender-receiver cells	Using protein-protein interactions at the interface of two synthetic cells to generate a signal.	Moghimianavval et al.83
Phospholipid vesicles ↔ Phospholipid vesicles	Glucose and luminol as external signals and light generated by HRP through oxidation of luminol as the juxtacrine signal	Photoactivation of iLID and adhesion of sender cell to receiver cell	Increase in the signal of a H2O2 indicator (roGFP-Orp1) in receiver cells and increased fluorescence of a lipophilic Ca2+ indicator in sender cells	Encoding bidirectional communication through H2O2 and Ca2+ upon chemiluminescence-mediated contact formation between sender and receiver cells	Ji et al.82
Water-in-oil emulsion droplet → water-in-oil emulsion droplet	3OC6-HSL, arabinose, or DFHBI	Activation of gene expression regulated by pLux or pBAD or binding of DFHBI to Spinach aptamer	GFP or RFP expression or improved fluorescence of Spinach aptamer	Demonstration of different kinetics of processing membrane permeable and impermeable signals. Using juxtacrine signalling for demonstration of feedforward and feedback motifs.	Dupin and Simmel78
Water-in-oil emulsion droplet ↔ water-in-oil emulsion droplet	3OC6-HSL from sender to receiver cell and 3OC12-HSL from receiver to sender cell	Activation of gene expression regulated by pLux and pLas promoters in receiver and sender cells, respectively	GFP expression	Implementation of two quorum sensing mechanisms separately in sender and receiver cells to demonstrate bidirectional juxtacrine signalling	Gonzales et al.79
Network of droplet interface bilayers	Light as the external signal and flow of ions between synthetic cells as the juxtacrine signal	Light-activated T7 promoter and electrical recording	Expression of membrane pore α-hemolysin to create conductivity between the juxtaposed synthetic cells	Demonstration of light-activated gene expression in DIB-based synthetic cells enabling long-range communication between juxtaposed synthetic cells within a synthetic tissue	Booth et al.54
Phospholipid vesicles → U2OS cells	Green fluorescent calcein dye or a NF-κB essential modulator (NEMO)-binding domain peptide (NBDp)	Gap junction formation through connexin 43 or connexin 32	Accumulation of calcein in the receiver cell cytosol or inhibition of luciferase expression regulated by NF-κB promoter in receiver cell	Cell-free protein synthesis of connexin 43 or connexin 32 upon formation of synthetic cells, enabling synthetic cell-natural cell signalling through connexon gap junctions	Kaneda et al.92

Autocrine signalling

Autocrine signalling in natural cells refers to self-signalling when cells secrete a signal for which they possess the receptor for (Figure 2a, left). One example of autocrine communication is signalling through growth factors and cytokines. When cells of the same type form a community, the secreted autocrine signalling molecule is sensed not only by the sender cell, but also by other adjacent cells. Relying on this mechanism, CD4⁺ T-cells use autocrine signalling via IL-2 to regulate their population size.⁵⁷ In the context of synthetic cells, autocrine signalling can be viewed as a mechanism in which the synthetic cell responds to a self-generated signal (Figure 2a, left). Cascades of chemical reactions processed by spatially separate compartments or subcellular synthetic organelles of a synthetic cell, thus, are examples of synthetic cell autocrine signalling. However, while this definition of autocrine signalling for synthetic cells includes sensing self-generated signals within the synthetic cell, it is functionally different from its natural counterpart. Natural cell autocrine signalling relies on signalling molecules secreted to the extracellular environment by the sender cell which could be sensed by neighboring cells, thus interfering with their autocrine signalling. However, the current examples of autocrine signalling in synthetic cells generates signals within the cell that can affect the neighboring synthetic cells only if the signal can traverse across the membrane and if the concentration of sender cells and the signal is high enough.

Figure 2∣. Autocrine and paracrine signalling in synthetic cells.

a. Schematics of autocrine signalling is shown in left. Natural cells secrete signals for which they possess the specific receptor. However, in synthetic cells, autocrine signalling often refers to cascades of reactions between compartments within the synthetic cell. Right: schematics of paracrine signalling mechanism. Sender cell releases diffusible molecules that elicit a response from the nearby receiver cell. b. Synthetic cell autocrine signalling mechanism consisting of cascades of reactions between compartments within the synthetic cell where the product of a reaction is the incoming autocrine signal of the other. The figure was redrawn based on ref. 35. In this autocrine signalling scheme, the entry of glucose into a synthetic cell triggers autocrine signalling between LUV organelles generating hydrogen peroxide and an internalized lipid vesicle oxidizing Amplex red and converting it to resorufin. c. Autocrine signalling between coacervate-based organelles within a droplet-based synthetic cell. The left panel shows the structure of DNA networks that construct the coacervates. The right panel depicts light-guided two-way signalling between coacervates relying on base-pairing of the signal and DNA sticky ends on each organelle. UV and visible (Vis) light determine the accessibility of the signal to sticky ends, thereby switching the direction of signalling. The figure was redrawn based on ref. 59. d. Paracrine signalling between synthetic cells occurs when diffusible molecules move from sender to receiver synthetic cells specialized for processing the signal. The schematics shows an example of paracrine signalling via ssDNA between proteinosome-based synthetic cells. An input activates signalling in sender cells by releasing the signal through toehold-mediated DNA strand displacement. The signal then diffuses to receiver cells and generates a programmed response. Signal amplification occurs when the receiver cell releases the signal and activates other receiver cells via diffusion. The figure was redrawn based on ref. 46.

An early demonstration of autocrine signalling in droplet-based synthetic cells was reported by Aufinger et al. by encapsulation of a cell-free protein synthesis system and agarose gel-based organelles in synthetic cells.⁵⁸ Their internal signalling mechanism used mRNA strands, encoding a fluorescent protein, controlled by a toehold switch and transcribed in an organelle specialized for transcription. The mRNA signal was then detected in a spatially separate organelle inside synthetic cells equipped with the complementary sequence of the toehold switch, thus enabling response generation to the self-produced signal by translation.

In a more cell-like design, GUV-based synthetic cells containing large unilamellar vesicles (LUVs) encapsulating glucose oxidase and micron-sized lipid compartments enclosing horse radish peroxidase and Amplex Red were generated by Li et al. (Figure 2b).³⁵ The membranes of both the synthetic cell and LUVs were permeabilized by melittin, a membrane pore. Autocrine signalling was then triggered by the addition of glucose to the synthetic cells, resulting in the generation of hydrogen peroxide by glucose oxidase in the LUVs. Finally, the autocrine signal (i.e., hydrogen peroxide) was detected inside the lipid compartments encapsulating horse radish peroxidase leading to the conversion of non-fluorescent Amplex Red to fluorescent resorufin.

A distinct difference between membrane-bound compartments and coacervate-based ones is the lack of a barrier between coacervates and their outer environment, which enables passive and rapid mixing of signals with the coacervate-based compartment. Leveraging this advantage, autocrine signalling mediated by single-stranded DNA (ssDNA) between DNA-based coacervate organelles inside a droplet-based synthetic cell was engineered by Zhao et al.⁵⁹ The DNA-based organelles were generated via base-pairing of sticky ends of DNA Y-motifs that generated large, ~10 μm, gel-like DNA networks (Figure 2c). The ssDNA signal was initially hybridized to an organelle containing an accessible complementary DNA sequence at the extension of its sticky end, while the extension of the sticky end of the other organelle containing the complementary DNA sequence of the ssDNA signal was blocked by a hairpin structure. In addition, the complementary sequence on both organelles was tethered to an azobenzene, making their ability to hybridize with the ssDNA signal dependent on the cis/trans conformation of azobenzene triggered by its photo-isomerization by Vis/UV light, respectively. Using this method, Zhao et al. showed autocrine signalling in synthetic cells in response to UV light through which ssDNA signal molecules were released from one organelle and detected and captured by the other. The inter-organelle communication inside the synthetic cell was also designed to be bidirectional, where exposure to visible light caused another cycle of autocrine signalling through which the signal was returned to its original organelle.

Similar to the examples mentioned above, the majority of other synthetic cell autocrine design schemes include cascades of reactions that are typically triggered by external stimuli and result in detectable fluorescent changes inside synthetic cells.^60,61 This is in stark contrast with natural autocrine signalling where important cell regulatory decisions such as growth and development are controlled and driven. Hence, encoding cell-like behaviors such as DNA replication and division as outputs of autocrine signalling mechanisms would be an important next step towards utilizing autocrine signalling for directing collective functions like controlling populations of synthetic cell communities.

Paracrine signalling

Short-range signalling via diffusible chemical signals is known as paracrine signalling (Figure 2a, right). Contrary to autocrine signalling, sender cells in paracrine signalling secrete molecules that only specific receiver cells, which harbor receptors for the signal, can sense. Paracrine signalling between cells through growth factors,⁶² neuropeptides,⁶³ hormones,⁶⁴ and even nitric oxide gas⁶⁵ controls processes such as cell proliferation, tissue repair, and cell metabolism. Paracrine signalling also plays critical roles in bacterial intra- and interspecies communication, driving processes including quorum sensing, virulence factor production, and swarming motility.⁶⁶

Paracrine signalling between porous synthetic cells encapsulating a nucleus-like clay crystal non-covalently bound to DNA was demonstrated by Niederholtmeyer et al.⁶⁷ The semi-permeable membrane of synthetic cells consisted of a photopolymerized resin composed of glycidyl methacrylate and trimethylolpropane ethoxylate triacrylate which allowed transport of molecules as large as 2 MDa across the membrane. Two populations of synthetic cells, containing either DNA sequences for signal generation or detection, were immersed in a solution containing a CFPS system. In this setting, sender cells encapsulated DNA strands encoding the signal, T3 RNA polymerase, which drove transcription of mRNA molecules that were translated inside and outside the synthetic cells. The receiver cells included DNA sequences of TetR-GFP under pT3 promoter, along with a TetR binding motif, TetO. Therefore, when the signalling protein freely diffused into the receiver cells, it promoted transcription and translation of TetR-GFP as the signal detection mechanism. The tight binding of TetR-GFP to TetO sequence in receiver cells subsequently demonstrated signal detection. Notably, if sender cells also contained the signal detection mechanism, the system displayed quorum sensing-like behavior where the signal detection was successful only when the density of sender/receiver cells was higher than a threshold.

A paracrine signalling system that used glucose as the paracrine signal was demonstrated by Tang et al.⁶⁸ Sender cells encapsulating glucose and a CFPS system could communicate with proteinosome receiver cells, which displayed glucose oxidase on their membrane and contained horseradish peroxidase. Sender cells contained a DNA template encoding the sequence of the membrane pore alpha-hemolysin downstream of an esaO operator. Therefore, their exposure to N-(3-oxohexanoyl)-L-homoserine lactone (3O6CHSL) activated the transcription-translation of the pore, inducing the release of glucose. Nearby receiver cells then converted the glucose to hydrogen peroxide via their membrane-tethered glucose oxidase. Lastly, the hydrogen peroxide was consumed by horseradish peroxidase to produce resorufin from Amplex Red present in the solution, generating a detectable fluorescent readout.

Booth and coworkers have explored light-based approaches to control intercellular communication between synthetic and living cells.⁵³ UV light-activated chemically modified DNA templates were used to spatiotemporally control gene expression in synthetic cells to produce and secrete membrane-permeable N-isovaleryl-L-homoserine lactone (IV-HSL). IV-HSL was taken up by neighboring E. coli receiver cells, which activated expression of a BjaR-regulated reporter gene. The development of light-activated synthetic cells opens the doors to precisely studying mechanisms of cellular communication in synthetic cell populations and with living cells.

While signalling among coacervate-based compartments has been mainly explored in the context of communication between coacervate organelles within synthetic cells, a recent study by Nikroo and colleagues demonstrated paracrine signalling between coacervate-based synthetic cells.⁶⁹ Coacervates were made by electrostatic association between positively charged quaternized amylose, negatively charged carboxymethylated amylose, and carboxymethylated amylose with a nitrilotriacetic acid (NTA) modification. The inclusion of NTA moiety allowed for entrapment of His-tagged protein signals inside the coacervate. Additionally, a His-tagged UV light-activated tobacco etch virus (TEV) protease was encapsulated in the coacervate, enabling light-mediated signal release through excision of the His-tag from the signal. The paracrine signalling design relied on a protein-based signal that consisted of a His-tagged GFP fused to a cRaf domain with double serine phosphorylation. The signal was initially bound to sender coacervates through its His-tag. However, after UV light exposure, the signal lost its His-tag and was released from sender cells, diffusing into receiver cells that included a cRaf binding partner, a dimeric 14-3-3 protein. The high affinity of phosphorylated cRaf to dimeric 14-3-3 proteins allowed receiver cells to ‘absorb’ the signal, demonstrating paracrine signalling between coacervate-based synthetic cells.

A series of creative works from the group of Tom de Greef pioneered paracrine signalling with DNA and RNA as orthogonal signalling molecules among proteinosome synthetic cells (Figure 2d).^46,70,71 The porous structure of proteinosomes allowed the free diffusion of short DNA and RNA strands into synthetic cells. This enabled the loading of signal generation, processing, and detection elements within synthetic cells, as well as initiating intercellular communication by the introduction of an input strand. The input and signal processing inside sender and receiver cells, respectively, were based on toehold-mediated DNA strand displacement. The signal detection was revealed by the dequenching of a reference strand in receiver cells due to the displacement of the quenching strand by the received signal. Utilizing this strategy, Joessar et al. demonstrated paracrine signalling between two and three communities of synthetic cells.⁴⁶ Additionally, they implemented a signal amplification scheme where the signal reception by receiver cells was followed by the displacement of the signal by an amplification strand, which was present in the solution but not accessible to receiver cells before signal detection. This strategy resembles the long-range ComQXP quorum sensing system in B. subtilis in which a signal can ‘awaken’ a cell and continue to propagate to other cells.⁷² Furthermore, the flexibility of DNA signal designs allowed the implementation of negative feedback communication, as well as distributed logic-circuit computation through intercellular communication between communities of synthetic cells.

Relying on the above-mentioned DNA strand signalling mechanism in proteinosomes, in a different work, Yang et al. introduced a photocleavable nitrobenzyl moiety to the quenching strand in sender cells, thereby making communication through ssDNA strands triggerable by external light irradiation.⁷⁰ Lastly, in a separate work, Yang et al. showed RNA-based signalling between porous proteinosomes.⁷¹ As opposed to previous designs, signalling was initiated by an input ssDNA, which after hybridization with a DNA strand in sender cells containing an incomplete T7 promoter would activate transcription of a single guide RNA (sgRNA). The resultant sgRNA, the signal, diffused outside to form a complex with a Cas protein, thus guiding it to the receiver cells encapsulating a DNA strand containing the PAM target sequence. The displacement of a quencher strand by a dCas9 or cleavage of a fluorophore-conjugated strand by Cas12a demonstrated signal detection inside receiver cells.

While paracrine signalling using transcriptional inducers or small metabolites constitutes the majority of studies on synthetic cell intercellular communication, novel schemes that use tailorable DNA and RNA molecules hold the potential for developing complex communication circuits encoding intricate, cell-like responses. Strategies of enzymatic or conditional signal amplification both extend the range of communication and increase signal detection sensitivity. Further, encoding distributed computation elegantly showcases the power of cooperation among communities of synthetic cells towards emergent function.

Juxtacrine signalling

Juxtacrine or contact-dependent signalling is a unique mode of signalling that occurs between two cells in physical contact (Figure 3a). The intercellular adhesion facilitates interactions between proteins residing on the surface of the sender cell with receptors on the receiver cells. The protein-protein interaction typically induces conformational or biochemical changes in the receptor proteins, thus driving downstream response pathways in the receiver cell. Developmental processes like tissue pattern formation through differentiation^73,74 and mature muscle generation via fusion of a founder cell with fusion-competent myoblasts⁷⁵ rely on juxtacrine signalling. In the former, the interaction of Notch-Delta proteins causes lateral induction or inhibition, thus driving pattern formation. Juxtacrine signalling enables targeted communication, thereby permitting specific and drastic responses. For instance, juxtacrine signalling allows immune cells to target only pathogens in a pool of cells and trigger apoptosis in receiver cells that they physically engage.^76,77

Figure 3∣. Synthetic cell juxtacrine signalling.

a. Juxtacrine signalling between cells that are physically in contact allows direct transfer of chemicals or protein-protein interactions at the contact interface of sender-receiver cells. b. Top: juxtacrine signalling between water-in-oil emulsion droplets connected through DIBs. A simple genetic circuit enables the implementation of an incoherent feedforward loop through diffusion of DFHBI from sender cells to adjacent receiver cells. The bottom shows pulse-like propagation of the fluorescent signal DFHBI from the signal source (leftmost droplet) down the chain of receiver cells. The figure was redrawn based on ref. 78. c. Design of a juxtacrine signalling mechanism between phospholipid vesicle-based synthetic cells. Renilla luciferase (RLuc) generates luminescence within synthetic cells upon the addition of its substrate, coelenterazine (CLZ), which induces adhesion between sender and receiver cells, thus allowing the flow of Ca²⁺ ions from the sender to the receiver cells. The receiver synthetic cell encapsulates Rhod2, a Ca²⁺ indicator, and Ca²⁺-activated PLA2 that lyses the receiver cell in response to the signal. The bottom shows the lysis of the receiver cell (red) when juxtaposed with two sender cells (green). The figure was redrawn based on ref. 81.

Despite its advantages, juxtacrine communication has been rarely implemented in vesicle-based synthetic cells, partly because adhesion and intermembrane protein functionalization in synthetic cells is a nascent development. However, virtually all communication schemes implemented for DIB-based synthetic cells can be considered juxtacrine because signal exchange between the compartments requires either direct or a chain of physical contacts.

Booth et al. showed juxtacrine signalling in a tissue made of juxtaposed aqueous droplets in oil, connected by DIBs.⁵⁴ Hundreds to thousands of synthetic cells, containing an in vitro transcription-translation system and DNA encoding alpha-hemolysin, were 3D-printed to form a ‘synthetic tissue’. Furthermore, the promoter region of the DNA was conjugated to a biotin moiety via a photocleavable linker. The binding of streptavidin to the DNA-bound biotin thus inhibited transcription in the absence of UV illumination. By exploiting patterned light illumination, alpha-hemolysin-mediated ion transfer was only detected through specific activated pathways of connected synthetic cells, measured by electrical recording.

Novel juxtacrine communication motifs in DIB-based synthetic cells were demonstrated by Dupin and Simmel.⁷⁸ In a network of connected water-in-oil synthetic cells, pulse-like signalling was established by implementing an incoherent feedforward loop. In this setting, the fluorophore 3,5-difluoro-4-hydroxybenzylidene imidazolinone (DFHBI), as a signal, diffused from a sender cell through multiple connected receiver cells (Figure 3b). Inside receiver cells, the DFHBI bound to encapsulated Spinach RNA aptamer, thereby increasing its fluorescence as a sign of signal detection. However, the DFHBI binding to Spinach also released a ssDNA strand which completed the promoter region of a transcriptional unit in receiver cells that generated a Spinach-sequestering RNA. This feedback design revealed a pulse-like signal sent from the sender cell to the receiver cells.

Using the same platform of water-in-oil droplets connected through DIBs, Gonzales and coworkers demonstrated bidirectional juxtacrine communication using Lux and Las quorum sensing machinery.⁷⁹ Sender cells initiated signalling by producing LuxI, using a CFPS system, thus synthesizing and sending 3OC6-HSL to the receiver cell. The receiver cell encapsulated LuxR and a plasmid encoding for LasI and GFP under the pLux promoter. Signalling from the sender cell, therefore, induced CFPS of GFP and LasI, followed by synthesis of 3OC12-HSL. Signalling then occurred in the opposite direction, via diffusion of 3OC12-HSL back to the sender cells, which activated LasR and CFPS of GFP under the control of a pLas promoter.

With the advances in optogenetics tools, several novel juxtacrine signalling mechanisms between vesicle-based synthetic cells have been established. By decorating the outer membrane of synthetic cells with a photoswitchable protein iLID and its binding partner, SspB, Chakraborty et al. demonstrated adhesion between synthetic cells triggered by external light.⁸⁰ In a separate study, Chakraborty et al., used the above adhesion mechanism to implement contact-dependent communication between sender cells encapsulating Ca²⁺ ions and a luciferase protein and receiver cells containing phospholipase A (PLA) (Figure 3c).⁸¹ PLA became enzymatically active through binding with Ca²⁺ ions and degraded the membrane of the receiver cell, mimicking apoptosis. By introducing the alpha-hemolysin pore and a pore blocker to the communities of synthetic cells, conditional transport of Ca²⁺ ions between sender and receiver cells could only be established once luciferase-dependent adhesion between synthetic cells occurred.

Using a similar iLID and SspB adhesion system as above, Ji et al.⁸² demonstrated bidirectional communication between juxtaposed synthetic cells by replacing the luciferase with a combination of glucose oxidase and horseradish peroxidase. In their work, the receiver cell contained a hydrogen peroxide sensor and Ca²⁺ ions, whereas the sender cells had a lipophilic calcium indicator on their membranes. Conversion of glucose to hydrogen peroxide, by glucose oxidase, initiated the reaction of luminol with horseradish peroxidase that generated chemiluminescence, thus driving adhesion between sender cells and receiver cells. After contact formation, the hydrogen peroxide diffused to the receiver cell and induced a blue shift in the excitation wavelength of the sensor as an indication of the signal reception. Communication in the opposite direction occurred through the diffusion of Ca²⁺ ions from receiver cells to the sender cells. The presence of Ca²⁺ ions in the sender cell was detected through the change in fluorescence of a membrane-bound calcium indicator.

Finally, recent work by Moghimianavval et al. demonstrated engineered juxtacrine signalling between synthetic cells through protein-protein interactions in the contact interface of synthetic cells.⁸³ To generate light as the signal, the fragments of a split luciferase protein were each fused to SpyTag and SpyCatcher, respectively. The resultant fusion proteins were linked to the outer membrane of either sender or receiver cells. In addition to harboring a fragment of the split luciferase, the receiver cells encapsulated iLID linked to their inner membrane and a cytosolic fluorescently labeled SspB. Since SpyTag and SpyCatcher have a high affinity for each other, their interaction in the membrane-membrane interface of sender-receiver cells drove split luciferase reconstitution and activation. The signal generated then activated iLID, thereby driving membrane recruitment of fluorescent SspB as the indication of signal reception.

Membrane receptors on receiver cells that induce conformational changes in their cytoplasmic domains after interacting with the signalling protein on the sender cells are an unmet need in synthetic cell juxtacrine signalling. These receptors not only make the signalling cascade more cell-like but will also enable versatile responses. In top-down synthetic biology, SynNotch is the gold standard for engineering tunable contact-dependent signalling.⁸⁴ SynNotch allows flexible ligand sensing and a wide range of cellular responses. Therefore, the development of a SynNotch-like platform for constructing juxtacrine signalling between synthetic cells holds great potential for engineering multicellular synthetic cell communities and synthetic tissues.

Applications of engineered synthetic cell communication systems

The recent burst of synthetic cell communication methods has driven the development of a diverse array of synthetic cell applications for specific purposes (Figure 4). The majority of research has focused on using synthetic cells to study natural processes, many of which are described above. Other applications seek to control living systems, using synthetic cells as targeted messengers to extend the natural communication range of cells, and to activate specific cellular pathways. The ability to communicate with natural cells also greatly lends synthetic cells to medical applications, in an emerging gap between simple molecular drugs and complex living cellular therapies.

Figure 4∣. Current and future applications of developing intercellular communication for synthetic cells.

a. GUVs signaled the differentiation of nearby neural cells through expressing a differentiation factor, which was released upon receipt of a signal which drove the co-expression of a protein pore.⁴⁰ This showed how synthetic cell communication is being applied to precisely study biological processes. This figure was redrawn based on ref. 40. b. Once injected into mouse tumour tissue, GUVs expressed and released an anti-cancer protein signaling apoptosis in neighbouring cells.³⁸ This highlights steps towards the development of targeted synthetic cell-based therapeutics. c. A representation of the current applications being explored for synthetic cells alongside those we envisage synthetic cells having a major contribution to in the future.

The simplicity and the level of control offered by synthetic cell-based communication platforms allow for a reductionist approach to interrogating biological mechanisms. For instance, morphogen gradients across a tissue during development cause differential gene expression in stem cells, driving their differentiation along specific trajectories. Implementation of a simple feedback genetic circuit along with juxtacrine signalling in a small network of DIB-based synthetic cells has demonstrated the mutual relationship between the positional information of neighboring cells in a tissue and their gene expression profile.⁸⁵

The ability to in situ synthesise and release signalling molecules enables the application of synthetic cells to modulate natural communication pathways, inducing and studying phenotypic changes more precisely than current methods. To this end, synthetic cells were programmed to constitutively express a mammalian differentiation factor: brain-derived neurotrophic factor (BDNF).⁴⁰ Upon detection of a natural biological signalling molecule, a homoserine lactone (HSL) signal, the synthetic cells then co-expressed the large bacterial pore-forming protein, perfringolysin O (PFO), to enable the release of BDNF. The released BDNF drove the differentiation of nearby neuronal stem cells (Figure 4a). Inspired by the ionic current generation from an electric eel, in a different study, a network of connected hydrogel droplets has been used to modulate the activity of neighbouring neuronal organoids.⁸⁶ This was achieved by positioning salt droplets alongside anion- and cation-selective droplets within the hydrogel chain, causing unidirectional movement of anions and cations. This current generation shows further potential for synthetic cells in energy applications. Through interacting with bacteria, synthetic cells also hold the potential to dynamically alter microbial communities, improving their stability and enhancing productivity. For instance, synthetic cells were used to sense theophylline, a signal that E. coli is insensitive to, and release IPTG, a signal that E. coli was engineered to respond to.⁸⁷ Synthetic cells have also been engineered to synthesise the 3OC12 HSL upon detection of the distinct 3OC6 HSL, both natural bacterial signalling molecules.³⁹ By incubating these synthetic cells with two non-communicating populations of bacteria, they mediated the interaction between the two, detecting 3OC12 HSL from Vibrio fischeri and sending 3OC6 HSL to engineered E. coli.

The ability to interface synthetic cells with living cells also holds great potential for developing smart medical devices. By leveraging the ability of synthetic cells to detect biological metabolites, biosensors can be prepared to monitor physiological functions. For example, luciferase-encapsulated, alginate-based synthetic cells were used to measure the activity of cocultured hepatocytes.⁸⁸ Released D-cysteine, from the synthetic cells, could be converted to D-luciferin by P450 1A2 secreted from the hepatocytes, a protein indicator of liver function. The encapsulated luciferase could then detect the extent of D-luciferin formation, and hence hepatic activity. Synthetic cells can also be engineered to autonomously sense and neutralise disease states. In this manner, synthetic cells that mimic pancreatic beta cells have been developed to sense high blood glucose and release insulin.⁷ A nested liposome design was used, where glucose was converted to a pH signal using glucose oxidase. This reduction in pH led to the fusion of internal insulin-containing vesicles with the outer lipid membrane, thus releasing insulin. Through directly interfacing synthetic cells with eukaryotic cells, there are emergent fields such as bottom-up synthetic immunology, which aims to create antigen-presenting synthetic cells⁸⁹ or bottom-up synthetic killer cells.⁹⁰

Synthetic cells also have the potential to synthesise and deliver therapeutic molecules at the site of disease, thereby reducing patient side effects by mitigating off-target drug toxicity. Gene-expressing synthetic cells containing DNA-encoding a potent anticancer bacterial toxin, Pseudomonas exotoxin A, have been injected into mice tumours.³⁸ Once inside the tumours, the synthetic cells expressed and released the toxin, inducing apoptosis of the neighbouring tumour cells (Figure 4b). Synthetic cells can also be implanted in target tissues. This could enhance therapies that require prolonged drug delivery, such as the treatment of chronic illnesses, and tissue engineering strategies. Synthetic cells, encapsulated within a Matrigel matrix plug, stimulated the formation of new blood vessels in mice by synthesising and releasing a recombinant version of the human basic fibroblast growth factor.⁹¹ Although not tested in an in vivo model, juxtacrine transfer of therapeutic peptides from synthetic cells to U2OS cells through connexon pores was demonstrated in vitro, indicating the potential of juxtacrine mechanism for targeted drug delivery.⁹²

Implications/future directions

From being capable of a basic task like inducible gene expression to controlling the differentiation of neurons, synthetic cells have transformed from simple droplets into powerful, smart cell mimics. However, it was only through emergent behaviors such as cooperation that natural cells could surpass the limits of a ‘single cell’ function. Analogously, engineering cooperative functions in synthetic cells to tackle environmental, sustainability, and biomedical challenges require not just a powerful synthetic cell, but a network of communicating ones. Division of labor is a feature ubiquitous in both primitive and advanced forms of life. Learning from nature, the next generation of synthetic cells will be equipped with adequate tools to perform biological functions as a part of a larger collective. For instance, a community composed of synthetic cells specialized for breaking down specific chemical compounds might send their products to a second synthetic cell population engineered to recycle energy from incoming signals, which could open a new paradigm of synthetic-cell-based approaches to address pressing environmental and sustainability challenges (Figure 4c).

As described earlier, many applications of synthetic cell-natural cell communication are limited to instances of chemical release by synthetic cells. Efforts in developing molecular tools, e.g., juxtacrine signalling proteins, to capacitate synthetic cells for specific one-to-one signalling with targets such as cancer cells or immune cells will transform our perspective towards biomedical applications of synthetic cells. We also envision hybrid communities of synthetic and natural cells where synthetic cells can monitor the environment and signal to natural cells for specific responses.

Overall, the trajectory of advances in synthetic cells promises not only the development of more cell-like functionalities but also the advent of emergent behavior and sophisticated forms of communication with natural cells that will expand the applications of synthetic cell communities in both basic and translational science and into the unknown.