Scientists carried out their experiments in the bacterium Escherichia coli. Cells of the species are artificially colored blue in this microscope image. IMAGE POINT FR / NIH / NIAID / Universal Images Group via Getty Images Almost all life on Earth builds its proteins from the same set of 20 building blocks, called amino acids. Some organisms have an extra one or two, but none have fewer. Now, researchers have used artificial intelligence to reengineer a bacterium so that some of its key machinery—not the entire microbe—can function with a missing amino acid.
The work, published in the journal Science on April 30, suggests that proteins can be trimmed and remain stable. The findings pave new avenues for research in the field of synthetic biology and allow scientists to understand what early life may have looked like.
“It’s a tremendous tour de force,” says Kaihang Wang, a Caltech synthetic biologist who was not involved in the study, to Elie Dolgin at Nature. But, he notes, “it’s a first baby step of a grand journey” to creating a cell that runs on a reduced set of amino acids.
For years, synthetic biologist Harris Wang, of Columbia University, had been curious about whether organisms could survive with fewer than 20 types of amino acids. Earth’s earliest life forms may have relied on a simpler set of building blocks, reports Jacek Krywko at Scientific American.
So, Wang and his colleagues first evaluated which amino acid could feasibly be removed from proteins. They chose isoleucine, because it can be replaced by similarly structured amino acids valine or leucine. Then, they introduced genes into the bacterium Escherichia coli that replaced isoleucine with valine or leucine in some of the most important proteins in the microbe’s ribosomes. These vital organelles act as tiny factories to produce proteins.
But that didn’t go so well. While the bacteria survived, only about 43 percent of them remained functional.
A.I. models, including Google DeepMind’s AlphaFold2, helped the team generate designs with the best replacements for the missing amino acid while preserving the proteins’ structures and functions. “Some of these A.I. designs were really surprising,” Wang tells Scientific American. “They didn’t look like anything we would have anticipated.”
In 2024, two A.I. researchers at Google DeepMind, Demis Hassabis and John Jumper, who helped develop AlphaFold2 won the Nobel Prize in Chemistry. They shared the award with David Baker, a biochemist at the University of Washington, who has worked on computational protein design.
The researchers validated each A.I.-suggested design until they had an isoleucine-free version of each of the 50-some ribosomal protein. In the end, they were able to combine 21 of them in a functional cell. The generated bacterial strain, called Ec19, kept the genetic changes throughout its evolution for more than 450 generations.
“It’s a fairly monumental undertaking to trim the alphabet of life down to 19 amino acids,” Christopher Snow, a protein engineer at Colorado State University who was not involved in the work, tells Catherine Offord at Science. Even though the team couldn’t combine all the lean ribosomal proteins in one cell, Snow says that the study is “very impressive” and helps “deepen the understanding of the design rules of life.”
Synthetic biologist Tom Ellis agrees. The work addresses “a really interesting question that’s fundamental to the origin of life on Earth,” says Ellis, of Imperial College London, who was not involved in the study, to Scientific American.
The finding “lends support to the idea that [early] life was probably just fine for a while with a smaller palette,” he tells Science. “Even for very large machines like the ribosome, you don’t necessarily need to paint with all 20 colors.”
Sara Hashemi is a science writer and fact-checker currently based in New York City. Her work has appeared in Sierra, The Body, Maisonneuve magazine and more.
