Life might not have needed all twenty amino acids to begin
For the first time in the history of biology, researchers have removed one of the twenty amino acids considered essential to all known life and watched a living organism continue to grow and reproduce. Working with E. coli in a collaboration spanning elite research institutions, scientists used artificial intelligence to redesign the bacterial ribosome itself — not individual proteins, but the very machinery that reads the code of life. The achievement does not merely expand the toolkit of synthetic biology; it quietly reopens the oldest question in science: what does life actually require in order to be?
- Biology's most foundational rule — that all known life depends on exactly twenty amino acids — has been broken in a laboratory for the first time.
- The challenge was immense: thousands of proteins rely on isoleucine, making its removal a puzzle too vast for human researchers to solve manually, requiring AI to scan and redesign at a scale no scientist could match alone.
- Rather than rewriting the genome protein by protein, the team made a decisive architectural bet — targeting the ribosome, the cell's master assembly machine — altering 382 components while leaving its core intact.
- Eighteen of fifty engineered strains survived and reproduced, slower than natural E. coli but undeniably alive, proving the approach is viable even if not yet optimized.
- The discovery now points toward organisms engineered for pharmaceuticals, manufacturing, and the resource-scarce environments of space — life redesigned not just to survive on Earth, but to function where Earth's biology cannot.
In a research laboratory, scientists have accomplished something biology long held to be impossible: they created living E. coli bacteria that survive, grow, and reproduce without isoleucine — one of the twenty amino acids that textbooks have always listed as essential to all life on Earth.
The feat required artificial intelligence. Rather than manually redesigning the thousands of proteins that depend on isoleucine, the team used machine learning models to predict protein structures and identify viable substitutions at a speed no human researcher could approach. But the deeper insight was architectural: instead of rewriting the genome one protein at a time, they focused on the ribosome — the cellular machine responsible for assembling all proteins — modifying 382 components tied to isoleucine recognition while preserving the ribosome's core function.
Fifty modified strains were produced. Eighteen survived. They grew more slowly than ordinary E. coli, but they divided and persisted — a meaningful distinction from merely tolerating the change. The experiment also speaks to evolutionary history: scientists have long theorized that early life may have operated with a simpler genetic code, using fewer amino acids before accumulating complexity over billions of years. This is the first time that theory has been tested in a living organism and found to hold.
The implications extend well beyond the petri dish. Organisms engineered to function with a reduced amino acid set could be designed to produce pharmaceuticals, serve specialized manufacturing roles, or operate in environments — like the surface of Mars — where biological resources are too scarce for conventional life. For now, the bacteria remain in the lab, slower and more fragile than their natural cousins. But a threshold has been crossed: life's fundamental toolkit has been reduced, and life continued anyway.
In a laboratory at one of three elite institutions—Columbia, MIT, or Harvard—researchers have done something that seemed theoretically possible but practically impossible: they created living bacteria that thrive without one of the twenty amino acids that have been the foundation of all known life on Earth.
The organism in question is E. coli, a bacterium so common in labs that it has become almost mundane. But these versions are not mundane. They are missing isoleucine, an amino acid that biology textbooks list as essential. The bacteria don't just survive this absence. They grow. They reproduce. They persist.
The breakthrough required a tool that didn't exist a decade ago: artificial intelligence trained to understand protein language. Rather than manually redesigning each of the thousands of proteins that depend on isoleucine, the research team—published this week in Science—used machine learning models to predict protein structures and identify which amino acids could substitute for the missing one. The AI systems scanned genetic possibilities at a speed no human researcher could match. But the team's real insight was architectural. Instead of rewriting the entire bacterial genome protein by protein, they focused on the ribosome itself, the cellular machine that assembles all proteins. They rewrote 382 components related to isoleucine recognition and function while keeping the ribosome's core machinery intact.
The experiment produced fifty distinct modified strains. Eighteen of them survived. Not all grew at the same rate as normal E. coli—the engineered versions moved more slowly through their reproductive cycles—but they moved. They divided. They lived.
This matters because it settles a question that has haunted evolutionary biology for decades. Scientists have long suspected that early life, billions of years ago, may have evolved using fewer amino acids than modern organisms require. The genetic code may have been simpler once. But suspicion is not proof. No one had ever removed an amino acid from a living thing and watched it continue to live. The distinction is crucial: adding new amino acids to expand the genetic code is one thing. Subtracting one of the original twenty is another entirely. It means rewriting the rules that govern how cells read DNA and build the proteins that make life possible.
The implications ripple outward. If bacteria can function with nineteen amino acids, then the boundaries of what life requires are not as fixed as we thought. That opens doors in synthetic biology—the engineering of organisms for specific purposes. Researchers could design bacteria that produce pharmaceuticals in controlled environments. They could create microbes optimized for manufacturing processes. And they could imagine organisms built for conditions where Earth's biology would fail: the thin atmosphere of Mars, the radiation of deep space, environments where biological resources are scarce and every molecule must be used with absolute efficiency.
The work also suggests something about life's origins. If simpler genetic systems are viable, then the path from chemistry to biology may have been more straightforward than we imagined. Life might not have needed all twenty amino acids to begin. It might have started with fewer and added complexity as it evolved.
For now, the engineered E. coli exist in petri dishes. They are slower than their natural cousins. They are not ready to be released into the world. But they represent a threshold crossed: the first time humans have subtracted from life's fundamental toolkit and watched life continue anyway. What comes next is the harder work—understanding why these bacteria can function at all, and whether that understanding can be applied to create organisms that don't just survive with less, but thrive.
Notable Quotes
Early life may have evolved using simpler genetic systems built from fewer amino acids— Research findings in Science
The Hearth Conversation Another angle on the story
Why does removing an amino acid matter more than adding one?
Because adding is expansion—you're giving an organism new tools. Removing is constraint. You're taking away something the textbooks say it needs to survive. If it survives anyway, you've learned that the textbooks were wrong about what's essential.
But the bacteria grew slower. Doesn't that mean they were struggling?
It does. But slower growth is not failure. It's adaptation. The bacteria found a way to do the same job with one fewer tool. That's the remarkable part.
How did the AI actually help? Couldn't researchers have done this manually?
Not in any reasonable timeframe. The ribosome alone has thousands of moving parts. Each one that touches isoleucine had to be redesigned. The AI didn't just speed up the work—it made the problem solvable at all. It could see patterns across millions of genetic possibilities that a human would never have the time to examine.
What happens if you try removing a second amino acid?
That's the question everyone's asking now. We don't know if you can stack these changes. The bacteria might need all nineteen of the remaining amino acids. Or there might be more room to simplify. That's the frontier.
Does this change how we think about life on other planets?
It suggests that life might not need Earth's exact toolkit to exist. If we're looking for life elsewhere, we might be looking for the wrong thing. Life could be built from simpler chemistry than we assumed.