All improvements are contained within the specific gene cluster
Each year, millions of tonnes of plastic accumulate in places where no natural process can reclaim them — a slow crisis that has long outpaced human solutions. Researchers at the National University of Singapore have answered with LySE, a platform that harnesses the ancient speed of viruses to accelerate the evolution of bacterial gene clusters, guiding microbes toward the complex biochemical work of breaking down plastic. Where previous tools could only reshape small fragments of genetic code, LySE operates at a scale five times larger, and does so with a precision that keeps improvements exactly where they are intended. It is, in essence, a way of teaching life to solve problems that life has never encountered before.
- Plastic pollution accumulates faster than any existing biological or chemical remedy can address it, and engineering bacteria to digest it has stalled because the genetic machinery required spans dozens of genes working in concert.
- Previous evolution platforms either handled only small gene clusters or produced 'cheater' bacteria that gamed the selection process without genuinely improving, making reliable progress toward complex traits nearly impossible.
- LySE exploits an intentionally error-prone viral replication system to flood target gene clusters with mutations while keeping the virus too weakened to spread uncontrollably, allowing researchers to alternate between generating variation and testing which variants actually work.
- In laboratory trials, five rounds of LySE-driven evolution produced a bacterial strain that grew 50.9 percent more efficiently on ethylene glycol alone, with every beneficial mutation confirmed and transferable to entirely new bacterial species.
- The platform requires only standard lab equipment and no specialized phage expertise, positioning it as a broadly accessible tool for pharmaceutical biosynthesis, pollutant degradation, and AI-designed carbon-capture pathways.
Every year, millions of tonnes of plastic settle into landfills and ocean floors with no natural exit. Scientists have long wondered whether bacteria could be engineered to break those polymers into useful chemical building blocks — but plastic digestion is not a single-gene problem. It demands an entire cluster of genes working in coordination, like tuning every machine on a factory floor simultaneously.
Researchers at the National University of Singapore have developed LySE, a platform that makes this kind of complex genetic engineering faster and more reliable than anything previously available. The system is built around bacteriophage T7, a virus that naturally infects E. coli and bursts open bacterial cells within minutes. The NUS team engineered it to carry a small ring of DNA — a phagemid — containing the gene cluster they want to improve.
The central innovation is deliberate imprecision. LySE uses a DNA-copying enzyme that introduces mutations roughly 160,000 times more frequently than a bacterium's own machinery. This high error rate also damages the phage's own DNA, weakening the virus so it cannot spread uncontrollably. By adjusting the ratio of phage to bacteria, researchers can toggle between a mutation phase, where new genetic variations accumulate, and a selection phase, where those variants are tested in normal bacteria to identify which ones genuinely perform better.
The system overcomes two persistent failures of earlier methods. Traditional phage-based evolution is limited to gene fragments of about 8,000 DNA letters and tends to produce cheater bacteria that survive selection without actually improving the target gene. LySE handles clusters up to 40 kilobases — five times larger — and confines all changes strictly to the target cluster, preventing unwanted mutations elsewhere in the genome.
Validation came in two forms. The team confirmed that improvements to antibiotic resistance held when genes were transferred to new bacteria, and they optimized a five-gene pathway for metabolizing ethylene glycol. After five rounds of evolution with progressively less glucose available, the best-performing strain produced 50.9 percent more biomass on ethylene glycol alone. Sequencing confirmed that LySE had modified both the regulatory switches and the protein-coding genes themselves.
Because all improvements remain within the specific gene cluster, optimized pathways can be transferred cleanly into entirely different bacterial species. The workflow requires only standard laboratory equipment, making the technology accessible without specialized phage expertise. Looking ahead, the team aims to apply LySE to entirely synthetic metabolic pathways — including AI-designed routes for CO2 capture — optimizing computationally conceived enzymes until they function efficiently inside living cells. A patent has been filed.
Every year, millions of tonnes of plastic settle into landfills and ocean floors with no natural way out. Scientists have begun asking whether bacteria might be engineered to eat it—to break those polymers down into useful chemical building blocks. The problem is that plastic digestion is not a single-gene task. It requires an entire cluster of genes working together, each one calibrated precisely, like tuning every machine on a factory floor at once rather than replacing a single part.
Researchers at the National University of Singapore have developed a new platform called LySE that makes this kind of complex genetic engineering faster and more reliable than existing methods. The breakthrough lies in how they exploit a virus called bacteriophage T7, which naturally infects E. coli bacteria. The virus replicates quickly and bursts open the bacterial cell within minutes. The NUS team engineered it so that when the phage makes new copies of itself, it also packages a small ring of DNA—called a phagemid—that carries the cluster of genes the researchers want to improve.
The key innovation is deliberate sloppiness. To generate many new versions of those target genes, the researchers use a specially engineered DNA-copying enzyme that is intentionally error-prone. A normal DNA polymerase works like a precise photocopier; this one is deliberately sloppy, introducing mutations at a rate about 160,000 times higher than a bacterium's own copying system. This seems counterintuitive, but the high error rate is what makes the system controllable. Because the enzyme is so careless, it also damages the phage's own DNA, weakening the virus so it cannot spread uncontrollably. It can only destroy bacteria when added in large numbers. By adjusting the ratio of phage to bacteria, researchers can toggle between a mutation phase, where target genes accumulate new variations and get packed into fresh phage particles, and a selection phase, where those mutated genes are tested in normal bacteria to see which ones actually work better.
PhD candidate Shujian Ong, who led much of the work, explained the approach: the engineered virus replicates rapidly and breaks open the bacterial cell within minutes, carrying with it the phagemid containing the genes to be improved. The system sidesteps two major problems that plague other continuous evolution methods. Traditional phage-based evolution can only handle small pieces of DNA—about 8,000 letters long. It also produces "cheaters," bacteria that mutate their own DNA in ways that trick the test and help them survive without actually improving the target gene. LySE handles gene clusters up to 40 kilobases, five times larger than the most common phage-based method, and keeps all improvements strictly within the target cluster, preventing unwanted mutations elsewhere in the genome.
The team validated LySE in two ways. First, they tested whether improvements to antibiotic resistance persisted when genes were moved into new bacteria, confirming that changes were genuinely built into the target genes. Second, they optimized a five-gene pathway that allows bacteria to use ethylene glycol for growth and energy. After five rounds of evolution, with less glucose available each time, the best-performing strain produced 50.9 percent more biomass using ethylene glycol as its sole food source. Genetic sequencing revealed that LySE had changed both regulatory regions—the switches that control how much a gene is turned on or off—and protein-coding genes themselves, with each helpful mutation confirmed by reintroducing it individually into fresh bacteria.
Assistant Professor Fredens, who oversaw the research, noted that without LySE, a bacterium's natural instinct is to mutate its entire genome to find ways to consume more plastic, but it struggles to find optimal solutions that way. LySE improves the target gene cluster dramatically without accumulating unwanted mutations in the rest of the bacterium's DNA. Because all improvements are contained within the specific gene cluster, the optimized pathway can be easily transferred into entirely different bacterial species.
The platform's capacity to handle larger gene clusters opens doors to applications that were previously impractical: optimizing biosynthetic pathways for pharmaceuticals, engineering microbes that break down environmental pollutants, and evolving entirely synthetic metabolic routes for carbon capture. The workflow requires only standard laboratory equipment and the mixing of phage lysates with cell cultures, making the technology accessible to labs without specialized phage biology expertise. A patent has been filed for LySE.
Looking forward, the team plans to apply the platform to systems that are entirely synthetic and new to nature. A key target is engineering synthetic CO2-fixing metabolic pathways—computationally designed routes that have never existed in the real world—and optimizing them so they actually function efficiently inside living cells. With LySE, researchers can take AI-designed enzymes and metabolic pathways and rapidly optimize them to work in practice. That is where the massive potential lies.
Citações Notáveis
Without LySE, a bacterium's natural instinct is to mutate its entire genome to find ways to consume more plastic, but it struggles to find optimal solutions that way.— Assistant Professor Fredens
With LySE, we can take AI-designed enzymes and metabolic pathways and rapidly optimize them to work in practice. That is where massive potential lies.— Assistant Professor Fredens
A Conversa do Hearth Outra perspectiva sobre a história
Why is engineering bacteria to eat plastic so much harder than, say, making a bacterium resistant to an antibiotic?
Because plastic digestion isn't one problem—it's five or ten problems happening at once. You need multiple genes working together, each one tuned to the right level. It's like trying to improve an entire assembly line instead of just one machine.
And the old methods couldn't handle that?
They could handle small pieces of DNA, but not the whole cluster. And they had a cheating problem—bacteria would find shortcuts that looked good on the test but didn't actually improve the thing you wanted them to do.
So LySE fixes both problems by making the virus deliberately bad at copying DNA?
Exactly. The sloppiness is the feature, not a bug. It creates so many mutations that the virus itself gets weakened. That keeps it from spreading out of control, and it keeps the bacteria from gaming the system.
How much bigger can LySE handle compared to the old methods?
Five times larger. The old phage-based methods max out around 8,000 DNA letters. LySE can handle up to 40,000. That opens up whole categories of work that were impossible before.
What happens after you evolve the bacteria in the lab? Can you actually use it?
Yes. Because all the improvements stay within the target gene cluster, you can move that optimized pathway into completely different bacteria. It's modular—you're not carrying along all the random mutations the bacterium accumulated elsewhere in its genome.
What's the real prize here—is it plastic, or something bigger?
Plastic is the proof of concept. The real prize is synthetic biology. They want to take pathways that AI designed—things that have never existed in nature—and use LySE to make them actually work inside living cells. That's where the potential gets massive.