A functional protein is a functional protein, whether it evolved yesterday or millions of years ago.
In Osaka, scientists have done something quietly extraordinary: they reached back millions of years into the evolutionary record and rebuilt proteins that no longer exist, not from fossils, but from the logic written into the proteins that do. By reconstructing ancient light-sensing rhodopsins and coaxing them to function inside living bacteria, researchers have transformed the study of protein evolution from inference into experiment — a reminder that the past is never entirely gone, only waiting to be read.
- A long-standing puzzle in evolutionary biology — how one ancient protein family diversified into hundreds of radically different functions — had resisted every standard analytical tool thrown at it.
- The stumbling block was the messy, variable regions of rhodopsin proteins that conventional sequence-alignment methods couldn't handle, leaving the evolutionary story perpetually out of reach.
- A team at the University of Osaka built a new pipeline called ConsistASR that embraces the insertions and deletions others ignored, allowing them to reconstruct what ancestral rhodopsins must have looked like before they diverged.
- They then synthesized those ancient sequences and inserted them into living bacteria — and the resurrected proteins folded, stabilized, and behaved exactly as evolutionary theory predicted they should.
- ConsistASR is now publicly available, meaning the same method that brought ancient light-sensors back to life can be aimed at virtually any protein family, turning deep evolutionary history into something measurable.
In a laboratory in Osaka, researchers have found a way to resurrect proteins that haven't existed for millions of years — not from amber or bone, but by reading the evolutionary record encoded in the proteins alive today. The work, published in ACS Omega, sheds new light on how a single ancient protein family gave rise to the hundreds of microbial rhodopsins we see now, each performing a strikingly different job.
Microbial rhodopsins are membrane-embedded proteins found across bacteria and archaea, and they are remarkably versatile — some pump ions, others sense light or chemical signals. How proteins sharing nearly identical core structures evolved such different functions has long puzzled biologists. The standard approach of aligning protein sequences kept failing because the regions that vary most between rhodopsin types made clean comparisons impossible.
Haruto Ishikawa and his team took a different path, building a method that specifically accounts for the insertions and deletions that accumulate in those variable regions over time. They used it to reconstruct ancestral versions of two major rhodopsin branches, then synthesized those ancient sequences and expressed them in living Escherichia coli to see whether the proteins would actually work.
They did. Both ancestral proteins folded into stable molecules. The reconstructed ancestral schizorhodopsin pumped protons just as its modern descendants do; the ancestral heliorhodopsin did not — consistent with what modern heliorhodopsins are known to do. The proteins absorbed light at the expected wavelengths, behaving precisely as evolutionary logic predicted.
The team has released their analytical pipeline, ConsistASR, for other researchers to use. The approach transforms protein evolution from something inferred through sequence comparisons into something you can synthesize, place under a spectrometer, and watch respond to light — a tool for testing the deep past by rebuilding it.
In a laboratory in Osaka, researchers have figured out how to resurrect proteins that haven't existed for millions of years—not by pulling them from amber or bone, but by reading the evolutionary record written into the proteins that survive today. The work, published in ACS Omega, offers a new way to understand how a single family of ancient proteins diversified into the hundreds of different forms we see in microbes now, each one doing something entirely different.
Microbial rhodopsins are proteins embedded in cell membranes. They are ancient and abundant, found across bacteria and archaea, and they do remarkably varied work: some pump ions across membranes, others sense light, still others respond to different wavelengths or chemical signals. The puzzle that has long occupied evolutionary biologists is how proteins with nearly identical core structures—all built around seven transmembrane domains—could have evolved such wildly different functions. The standard tools for tracing protein ancestry, which align sequences side by side to find common patterns, kept failing on rhodopsins because the parts of the protein that stick outside the membrane varied so much from one type to another.
Haruto Ishikawa and his team at the University of Osaka took a different approach. Instead of ignoring the messy bits, they built a method that specifically accounts for the insertions and deletions that accumulate in those variable regions over evolutionary time. Using this technique, they analyzed two major branches of the rhodopsin family—schizorhodopsins and heliorhodopsins—and worked backward to reconstruct what the ancestral versions of each must have looked like. Then they did something remarkable: they synthesized those ancient sequences and inserted them into living bacteria to see if the proteins would actually fold and function.
They did. Both ancestral proteins produced stable, mature molecules in Escherichia coli. The ancestral schizorhodopsin showed the same light-driven proton-pumping activity as its modern descendants. The ancestral heliorhodopsin, by contrast, did not pump ions—consistent with what modern heliorhodopsins do. The proteins had distinctive colors and showed the characteristic light-absorption properties you would expect from functional rhodopsins. In other words, the reconstructed ancient proteins behaved exactly as the evolutionary logic predicted they should.
What makes this work significant is not just that it worked, but that it opens a door. The researchers have released their analytical pipeline, called ConsistASR, for other scientists to use. The same approach that brought ancient rhodopsins back to life could be applied to other protein families, other evolutionary questions. It is a tool for reading the deep past and testing whether your reading is correct by actually building what you think you found. The method transforms protein evolution from something you can only infer from sequence comparisons into something you can hold in your hand, measure in a spectrometer, and watch respond to light.
Citações Notáveis
Rhodopsins all have seven transmembrane domains that are very similar, but their extramembrane domains vary dramatically, making standard sequence alignment techniques insufficient to trace their evolution.— Haruto Ishikawa, lead author
Both ancestral sequences produced stable, mature proteins in bacteria that had distinctive color and showed characteristic spectral properties, just like existing rhodopsins.— Yasuhisa Mizutani, senior author
A Conversa do Hearth Outra perspectiva sobre a história
Why does it matter that they could make these ancient proteins work? Couldn't they just read the sequences and understand the evolution that way?
Reading sequences tells you what changed, but not what those changes actually meant functionally. A protein might accumulate mutations that look neutral on paper but completely alter what it does. By making the ancestral protein and testing it, they're checking whether their evolutionary story is correct.
So they're using the present to validate the past.
Exactly. They're saying: if our reconstruction is right, this ancient protein should fold properly and show these specific properties. And it did. That's powerful confirmation.
What's the practical application? Are they trying to engineer new proteins?
That's one possibility. But more immediately, they're building a toolkit. ConsistASR can be applied to any protein family where the standard methods fail. It's a way of asking evolutionary questions that were previously unanswerable.
And the bacteria—they just... accepted these ancient proteins?
The bacteria didn't know the difference. A functional protein is a functional protein, whether it evolved yesterday or millions of years ago. That's what made the result so striking. The ancient machinery still works.