Scientists use AI to create bacteria that run key cellular machinery on just 19 amino acids

All life runs on 20 amino acids. These cells run key machinery on just 19

All life on Earth depends on the same molecular alphabet: 20 amino acids that cells string together to make proteins. But now, scientists have reengineered bacteria to run a core part of their cellular machinery with just 19 of those amino acids -- a feat akin to rewriting one act of a Shakespearean play without a common letter like "R" while keeping the text intelligible. The work is reported today in Science. "It's very exciting that it's possible," says Julius Fredens, a synthetic biologist at the National University of Singapore who was not involved in the research. The work offers a blueprint for engineering cells with capabilities beyond those found in nature, the authors say, while also hinting at a simpler past when early life relied on a more limited set of building blocks. Researchers have long sought to rewrite the genetic code of life, both to expand what cells can do and to probe the basic rules of life. For example, scientists have streamlined DNA by removing sequences that encode the same amino acid as other stretches. But most researchers have left the 'canonical' 20 amino acids untouched, because even small changes to a protein's amino-acid sequence tend to disrupt its function. The challenge of subtracting a letter from the vocabulary of proteins intrigued Harris Wang -- though his early attempts fell short. A synthetic biologist at Columbia University in New York City, Wang initially tried simply swapping one amino acid -- isoleucine -- with others that differ slightly in size and shape, but fewer than half of his modified proteins remaining functional. Wang shelved the project for a few years, until a new generation of artificial-intelligence tools began to change what was possible. Systems such as AlphaFold can predict a protein's 3D structure, and various protein language models can now suggest entirely new amino-acid sequences that fold and function. Crucially for Wang, these tools could point to non-intuitive ways that might allow replacement of isoleucine without undermining protein performance. Still, reworking all 4,000-plus proteins in the bacterium Escherichia coli seemed too daunting a challenge. Instead, Wang chose a more focused, if still ambitious, target: the ribosome. The ribosome -- a complex of more than 50 proteins and catalytic RNA -- sits at the heart of the cell, translating genetic instructions into proteins. If such a critical system could operate without isoleucine, Wang reasoned, the same approach might extend to the rest of the proteome. Wang teamed up with computational biologists Sergey Ovchinnikov and Simon Kozlov at the Massachusetts Institute of Technology in Cambridge. The team leveraged AI models informed by evolution and protein structure to propose substitutions for isoleucine. The researchers experimentally validated the predicted designs, engineering bacteria to produce the rejiggered proteins and checking how well the cells grew, then repeating the cycle as needed until they arrived at a working, isoleucine-less version of each ribosomal protein. Most proteins proved amenable to this AI-guided design-and-test process. But a few needed brute-force, hands-on work in the laboratory. This mix of necessary methods highlights both the power and the limits of AI-driven protein engineering, notes Kozlov. "We have not solved biology yet with AI," he says. But compared to what was possible before, "this is a dramatic acceleration". In the end, the team successfully replaced all 382 isoleucines across the ribosome's various protein components. The authors generated a single E. coli bacterium that contained a subset of these altered proteins. That cell grew robustly, with only a minor slowdown compared to unmodified cells, and remained genetically stable for more than 450 generations. "It's a tremendous tour de force," says Kaihang Wang, a synthetic biologist at the California Institute of Technology in Pasadena who was not involved in the research. But, he adds, "it's a first baby step of a grand journey" toward building an entire cell that runs on the same 19-amino acid alphabet. Reaching that milestone would push the limits of life's chemistry, and open the door to new kinds of synthetic organisms, he says. Rapid technological advances should help accelerate progress toward that goal. But, as Caltech's Wang notes in an accompanying Science commentary, getting there will likely continue to require a mix of advanced computational design and human ingenuity. "The AI is tremendously powerful," he says. "But human input is still critical -- at least for now."

AI helps create bacterium that's partially missing a universal amino acid

Of the hundreds of types of amino acids found on Earth, it's a mystery why life settled on 20 as the building blocks for all its proteins. Although certain species can use more -- some microbes employ up to 22 -- no one's ever found one using fewer. But now scientists are closer to creating such an organism, after partially eliminating one of the 20 amino acids from the bacterium Escherichia coli. The research, published today in Science, used artificial intelligence (AI) to propose alternatives to the amino acid isoleucine in dozens of proteins making up bacterial ribosomes -- the protein factories of the cell. The findings offer a glimpse into how earlier, simpler life forms might have lived and suggest new ways to synthesize proteins with bespoke functions in medicine and biotechnology. "It's a really bold, fun idea," says Tom Ellis, a researcher in synthetic genomes engineering at Imperial College London who was not involved in the work. The modified bacteria may not have immediate applications themselves, but the findings showcase AI's ability to predict protein structures and could make it easier to create designer proteins in the future, he adds. In previous work, researchers have edited the genetic code of bacteria, yeast, and other organisms to incorporate additional amino acids, whether naturally occurring or synthetic. (Some pharmaceutical compounds, such as the weight loss and diabetes drug semaglutide, are also made with so-called noncanonical amino acids that increase a compound's stability, for example.) But Harris Wang, a systems and synthetic biologist at Columbia University, has long wondered whether organisms could get away with fewer building blocks -- after all, the earliest forms of life couldn't have sprung into being with the whole set. To evaluate which block might be the most expendable, the team studied how often each amino acid is naturally replaced by a different one in proteins across multiple types of bacteria. Isoleucine was frequently substituted, they found, usually with a structurally similar amino acid such as valine or leucine. Feeling ambitious, the team turned its sights to the ribosome. Eliminating an amino acid entirely from this machinery is "almost the hardest thing you could think about, because it's the biggest, most complicated protein complex" in the cell, Wang says. The researchers first attempted to cut isoleucine from each of the 50 protein subunits in the bacterial ribosome that contains the amino acid. They created various strains of E. coli, each with a different gene edited to replace any isoleucine-encoding DNA sequences with ones coding for valine. For more than one-third of the modified proteins -- 18 out of 50 -- that approach worked, and the resulting bacterial strains grew normally. AI tools such as , which predict protein structures from amino acid sequences, or vice versa, offered ways to compensate for swaps in the other 32 subunits and recover original protein structures. The computational tools proposed switching out amino acids neighboring the replaced isoleucine, for example, or substituting a distant amino acid that interacts with the edited site in the folded-up protein. Still, in a few cases, Wang's team had to resort to brute force, replacing each of a protein's amino acids until researchers found a combination that restored bacterial growth. Having successfully altered each protein individually, the researchers tried combining 21 of the rewritten ribosomal proteins in one bacterium. It didn't work right away, but with further tweaks, they managed to create a strain that could grow, albeit relatively slowly. "It's a fairly monumental undertaking to trim the alphabet of life down to 19 amino acids," says Christopher Snow, a protein engineer at Colorado State University who was not involved in the work. Although the team didn't achieve it for the whole ribosome simultaneously, the study is nevertheless "very impressive" and helps "deepen the understanding of the design rules of life." Wang says his team is excited to apply the approach across the rest of the E. coli genome -- and perhaps attempt a bacterium with just 18 amino acids. Organisms with a reduced dependence on particular amino acids might better survive hostile environments or resist infections by viruses, he speculates. Removing an amino acid entirely also "frees up" the specific DNA sequences that typically code for it -- so those sequences could be reassigned to encode other, perhaps synthetic amino acids to create new drugs or other molecules, Ellis adds. Intriguingly, the team's changes to E. coli seemed to be relatively stable, Snow notes: After 450 generations, the modified E. coli didn't revert to using isoleucine in the modified proteins. That finding "lends support to the idea that [early] life was probably just fine for a while with a smaller palette," he says. "Even for very large machines like the ribosome, you don't necessarily need to paint with all 20 colors."

Scientists use AI to test whether life can run on only 19 amino acids

An engineered E. coli strain survived after one amino acid was designed out of many of its ribosomal proteins -- an early test of whether life's chemistry can be simplified Nearly all known life builds proteins from the same alphabet of 20 canonical amino acids. Strung together in different orders, those building blocks form the proteins that make cells work. In a new Science study, researchers at Columbia University, the Massachusetts Institute of Technology and Harvard University used artificial-intelligence-guided protein design to test how much of that alphabet can be pared back: they engineered an Escherichia coli strain that survived after it was redesigned to not have a specific amino acid in its ribosomal proteins. The team did not create a true 19-amino-acid organism. The engineered strain still uses the targeted amino acid, isoleucine, throughout most of its genome. But the result suggests that one of life's most ancient and essential machines can tolerate at least partial simplification -- and that AI may help biologists test the limits of life's chemistry. "The underlying question that we seek to ask is what early life looks like," says Harris H. Wang, a professor of systems biology at the Columbia University Irving Medical Center and senior author of the study. Researchers think all life today descends from an ancient, single-celled organism that lived more than four billion years ago. But some suspect that earlier, simpler life-forms that predate even this common ancestor may have run on a leaner chemistry. Wang's team wanted to find out whether modern cells could be engineered in that direction. If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today. "Think about language. There are 26 letters in the English alphabet, but do you really need 26, or can you simplify that to 25 or 24?" Wang says. The team chose to remove isoleucine because it resembles the amino acids valine and leucine closely enough that, in principle, some proteins might tolerate isoleucine's removal when it was replaced with one of them. They worked with E. coli, one of biology's best-studied organisms, and targeted its ribosomes, the molecular machinery that builds proteins and is itself a sprawling complex of more than 50 proteins. "Like in a video game, we just pushed the 'skip to the final boss' button," Wang says. The first attempt was brute force. The researchers took 39 essential or highly expressed E. coli genes and replaced every isoleucine with valine or leucine, like a genetic find-and-replace. The engineered bacteria survived but did so poorly. Their fitness dropped to about 40 percent of wild-type E. coli. The team's target was 90 percent. To close the gap, the researchers turned to AI. They combined two kinds of models. First, sequence-based protein language models such as ESM2 and MSA Transformer read protein sequences and suggested evolutionarily plausible mutations that a simple swap would miss. Then structure-based AI models such as AlphaFold2 and ProteinMPNN checked that the redesigned proteins would fold into the correct shapes and fit alongside neighboring molecules. The proposals were stranger than the team expected. "Some of these AI designs were really surprising," Wang says. "They didn't look like anything we would have anticipated." In one case, while redesigning a ribosomal protein called RpsJ, the AI remodeled an alpha helix -- a structural element bridging different parts of the ribosome -- and introduced eight new nearby mutations to compensate for the substitution of just two isoleucines. "Maybe these machine-learning systems know some aspects of biology we can experimentally verify but we don't yet understand," Wang says. "A noteworthy part of the project is the evolving contribution of AI to this work," says Tom Ellis, a professor of synthetic genome engineering at Imperial College London, who was not involved in the study. "In the last seven years, the AI-enabled modeling of proteins and mutations in proteins has come on leaps and bounds." The team first tested each AI-suggested change one at a time, confirming individual edits could meet the 90 percent fitness goal. Combined, the changes killed the cells. So the researchers debugged the genome by hand. Starting fresh from the natural E. coli sequence, they added the AI-designed pieces in small batches until the cells stopped growing, narrowing down the lethal interaction to a single region so they could fix it. The final strain, Ec19, carries 21 isoleucine-free ribosomal proteins out of 52, alongside AI-redesigned versions of the others that the team validated individually but could not yet combine. The strain is robust: fitness stays above 90 percent of wild-type E. coli, and natural selection did not revert the changes over 450 generations. "The paper is a tour de force of synthetic biology to address a really interesting question that's fundamental to the origin of life on Earth," Ellis says. He adds that this work could eventually inform biotechnology beyond Earth, in environments where not every amino acid is available. For now, Ec19 remains a 20-amino-acid organism. Wang and his colleagues purged 382 isoleucine residues from ribosomal proteins, but the rest of its genome still contains more than 81,000 isoleucine residues across thousands of other proteins. A truly 19-amino-acid organism will require cheaper, faster DNA synthesis and more capable AI models, including genomic language models trained on whole genomes rather than just proteins. Still, showing that ribosomal proteins can survive even partial simplification gives researchers a template for the rest of E. coli. "Considering the ribosome is probably the oldest remnant of the original common ancestor organism that first evolved protein synthesis, it's also a poetic thing to demonstrate this ambitious work on," Ellis says.