Scientists have created tiny DNA droplets that behave like living cells: they grow, they shrink, they respond to stimuli, and they eventually disappear. The trick is that these synthetic microdroplets can be programmed to follow precise life cycles, controlled by enzymes and light. The work offers a new blueprint for creating artificial cells with programmable, dynamic behaviors that mimic real biology.

For decades, researchers have dreamed of building synthetic cells from scratch. Such creations could reveal fundamental secrets about how life organizes itself. They might also become sophisticated drug delivery vehicles or biological computers. But one persistent challenge has eluded solution: creating synthetic cells that behave transiently, that form and dissolve on command the way real cellular structures do throughout living organisms.

Real cells are not static machines. Proteins clump together, form temporary compartments, exchange molecules, and then disperse. These dynamic processes are crucial. They allow cells to respond, adapt, and survive. Building synthetic systems that mimic this transient behavior has proven remarkably difficult.

Now, researchers have demonstrated a solution using DNA itself as the building material. The system relies on a principle that nature has perfected: the controlled assembly and disassembly of molecular structures. By engineering DNA strands to be cleaved at precise locations by molecular scissors, the team has created synthetic microdroplets that grow on command and dissolve predictably.

How the System Works

The foundation of this approach is deceptively simple. The researchers began with Y-shaped DNA frameworks, three-armed structures built by carefully annealing strands of DNA. These Y shapes have exposed "toehold" sequences at the tip of each arm, like Velcro waiting for a match.

When complementary DNA strands are added, these toeholds bind to the Y-shaped units, cross-linking them together in three dimensions. The linked structures naturally phase separate, clustering together to form liquid droplets filled with DNA. Within minutes, green fluorescent droplets become visible under a microscope, growing steadily larger.

But here is the crucial innovation: the cross-linking strands were engineered to contain recognition sites for specific molecular scissors. When the enzyme EcoRI was added to the solution, it began cleaving the cross-linking strands. As the bridges were cut, the three-dimensional network loosened. The droplets that had been growing now began to shrink. Within three hours, they disappeared entirely.

The process is entirely reversible in principle, though in this system it runs to completion. By controlling the concentration of enzymes and DNA fuel strands, the researchers could adjust when droplets reached their maximum size and how quickly they dissolved. Higher enzyme concentrations accelerated depletion. More fuel strands meant larger peak sizes before collapse.

Multiple Systems, Multiple Controls

The beauty of DNA as a programmable material is its modularity. The team created multiple independent systems, each responsive to different enzymes. One system responded to EcoRI, another to HindIII, a third to nickases (slightly different molecular scissors that nick rather than cut both strands of DNA).

By mixing different DNA frameworks and fuel strands, the researchers could create cocktails where multiple types of droplets assembled simultaneously in the same solution. Green fluorescent droplets and red fluorescent droplets grew side by side, their sizes determined independently by which enzymes were present.

They then went further, using inhibitor strands to selectively block certain DNA modules. When an inhibitor bound to a Y-shaped framework, it prevented that structure from participating in droplet formation. This allowed the researchers to create "gated" systems: selectively turning off assembly of one droplet type while leaving another active. The possibility to control which synthetic structures form in a mixed population opens doors to building more complex multicellular-like systems.

Building Catalytic Droplets

The next leap was architectural. Rather than relying on external enzymes floating in solution, the team embedded catalytic units directly into the DNA droplets themselves.

They engineered Y-shaped DNA structures to incorporate Mg2+ dependent DNAzymes, enzymes made entirely of RNA and DNA rather than proteins. These catalytic DNA units were woven into the cross-linking strands, so whenever a droplet assembled, the DNAzymes self-organized into a functional network within the droplet itself.

The droplets were then presented with a hairpin-shaped DNA substrate containing a ribonucleotide. The embedded DNAzymes recognized and cleaved this hairpin, splitting it into fragments. One of those fragments was engineered to displace the cross-linking strands holding the droplet together.

The beauty of this system is that it creates a self-contained cycle. The droplets grow, organize their internal catalysts, and then systematically self-destruct by cleaving the very molecules that hold them together. The product of catalysis becomes the seed of the droplet's dissolution. Peak sizes were determined by the concentration of substrate molecules. More substrate meant slower decay.

The researchers confirmed the importance of the catalytic function by replacing the normal hairpin substrate with a modified version containing only DNA. When that all-DNA version was used, it could not be cleaved by the DNAzymes. The droplets assembled normally but persisted indefinitely, trapped in a static state. Only the catalytic cleavage event could trigger dissipation.

Light as a Control Switch

To add another layer of programmability, the team incorporated light-responsive sequences. Hairpin structures containing a photocaged group (a chemical group that blocks binding until exposed to UV light) were added to the system.

When the solution was illuminated at 365 nanometers, the photocaged hairpin uncaged, releasing a strand that could serve as fuel for droplet formation. The length of illumination determined how much fuel was released, directly controlling the subsequent growth rate.

More remarkably, the researchers could pulse light repeatedly. After initial illumination triggered droplet assembly and growth, the droplets would begin to decay as their internal DNAzymes cleaved the substrate. But if a second, shorter light pulse was applied hours later, more fuel was released, triggering regrowth. The droplets expanded again before beginning to shrink once more.

This creates a temporal, light-modulated cycle of growth and decay. The droplets responded to light stimuli by changing their morphology, expanding and contracting under photonic control. The researchers note that these light-triggered shape transitions emulate dynamic shape changes that real cells undergo during metabolic cycles.

Why This Matters

Current models of synthetic cells often create structures that are biochemically inert or static. They assemble but do not respond dynamically. These DNA-based microdroplets demonstrate something fundamentally different: programmable transience, the ability for a synthetic system to follow a predetermined temporal program without living cellular machinery.

The practical implications are substantial. Droplets engineered to release cargo over time could become sophisticated drug delivery systems, dispensing therapeutic molecules in controlled pulses. The timing and dose could be set by adjusting enzyme concentrations or fuel molecule quantities.

More broadly, the work shows that DNA can serve as a programmable scaffold for building artificial organelle-like compartments with tunable lifespans and behaviors. The modular nature of the approach suggests that more complex multicellular-like assemblies might be engineered by combining multiple responsive droplet types, each with distinct roles and temporal dynamics.

The researchers note that their method could be extended to create pH responsive microdroplets, or those controlled by G-quadruplex formation or by molecular recognition of small RNA molecules. The toolkit is not fixed but expandable.

Perhaps most fundamentally, these systems demonstrate that biology's most essential property, temporal organization and change, is not exclusive to living cells. It can be engineered from first principles using rational design of molecular interactions. The droplets are not alive, but they behave in ways that make living systems possible: growth, metabolism through molecular transformation, responsive behavior, and programmed demise.

This research suggests that the path toward artificial cells need not recapitulate the full complexity of genomic machinery. Instead, by carefully programming how molecules interact, assemble, and dissociate, synthetic systems can acquire the dynamism and responsiveness that characterize life itself.

Credit & Disclaimer: This article is a popular science summary written to make peer-reviewed research accessible to a broad audience. All scientific facts, findings, and conclusions presented here are drawn directly and accurately from the original research paper. Readers are strongly encouraged to consult the full research article for complete data, methodologies, and scientific detail. The article can be accessed through https://doi.org/10.1021/jacs.5c00637