Some of the most elegant machinery in biology consists of proteins that share a common origin yet have settled into very different jobs. How that divergence happens is a central question in molecular evolution. A team working across Kiel University, the German Electron Synchrotron DESY and the Center for Structural Systems Biology (CSSB) has approached that question by pressing rewind. Led by Holger Sondermann, professor at Kiel University and head of the Structural Microbiology group at DESY, the researchers reconstructed the common ancestor of two bacterial enzymes, produced the long-extinct protein in the laboratory, and used it to trace how distinct functions emerged over evolutionary time. Their findings, published in Science Advances, offer a rare glimpse into the molecular past and demonstrate a powerful strategy for studying protein evolution.
Two Enzymes, Diverging Roles
The two enzymes examined in the study belong to a group of proteins known as nucleases, which break down molecules built from the chemical units of genetic material. Although the two enzymes are related, they have come to occupy different niches inside the bacterial cell.
The first enzyme, NrnC, is a generalist of sorts. It breaks down small molecules consisting of two linked DNA or RNA building blocks. The second enzyme, diDNase, performs a similar reaction but has become specialized: it is mainly dedicated to processing molecules made up of two DNA building blocks. The difference may sound subtle, but for a bacterium the distinction matters. Being able to degrade one type of small nucleic acid but not another can affect gene regulation, DNA repair, and the recycling of genetic material.
The scientific puzzle, as Sondermann and his colleagues framed it, is not just what these enzymes do today. It is how such closely related proteins developed such distinct preferences. In the grand sweep of evolution, a single amino acid change can alter an enzyme's shape, stability, or activity. But tracing that path is difficult when only modern descendants are visible. The past must be inferred, and the most direct way to test those inferences is to recreate it.
The Limits of Studying Living Enzymes
The new study builds on several years of research by the Sondermann group. The team began in a conventional way: comparing proteins from different bacteria to determine which characteristics have remained conserved over time and which have changed. Conserved regions are often essential for structure or function, while variable regions may account for differences in specificity.
In earlier studies of a particular enzyme family, Sofia Mortensen, a scientist in the Sondermann group, identified a group of proteins that were closely related to known enzymes but processed different molecules. That discovery opened the door to investigating how functional divergence occurs within a family where the evolutionary relationships were already clear. The proteins were evidently siblings, yet they had grown apart in their chemical preferences.
The team initially tried to explain the differences between the two enzymes by introducing targeted changes into the proteins. This is a common approach in protein engineering: mutate one amino acid at a time and see which substitution flips an enzyme from one activity to another. But in this case, the strategy failed. The modified proteins either lost their stability or failed to perform the expected function. Tinkering with the visible surface of the enzymes was not enough to reveal the underlying path of evolution.
Looking Backward, Not Forward
Because present-day proteins resisted explanation, the researchers decided to take a different route. Rather than simply comparing the proteins that exist today, they looked back at their evolutionary history. This is the logic behind ancestral sequence reconstruction, a method that uses the sequences of many related proteins to infer the most probable sequence of a protein that existed millions of years ago.
The team first built a detailed evolutionary tree of the enzyme family, placing NrnC, diDNase, and other relatives according to how similar their amino acid sequences are. With that tree in hand, they were able to calculate the most likely sequence of the common ancestor from which both enzymes descended. Then came the truly ambitious step: they synthesized the gene, expressed it in the laboratory, and produced the ancient protein in a form that could be studied directly.
This resurrected protein is more than a relic. It serves as a starting point for experiments. By re-creating an ancestral state, the team can introduce changes one by one and observe how each change nudges the protein toward one modern function or the other. In effect, the ancient enzyme is a molecular time machine, allowing the researchers to replay evolution in a test tube.
A Structural View of the Molecular Past
The study also provides a structural perspective. Researchers were able to look at the 3D structure of the reconstructed ancestor of the diDNase enzymes. Seeing the shape of an ancient protein helps explain why certain mutations were tolerated and why others were not. It also clarifies how the active site, the region where the enzyme carries out its chemical reaction, was remodeled over time to accept different substrates.
Because the ancestral enzyme predates the split between NrnC and diDNase, its structure likely represents a state in which both functions were still possible. By comparing this ancestral structure with the modern structures of NrnC and diDNase, the team can identify the precise changes that tilted the enzyme toward one specialization. This kind of information is difficult to obtain by examining modern enzymes alone, because millions of years of subsequent evolution can obscure the early steps that set a protein down a particular path.
Implications for Fundamental Evolution and Protein Engineering
The findings provide insights into how proteins can evolve different functions over the course of evolution, but the implications extend beyond basic curiosity. Understanding how nature reshapes proteins is valuable for scientists who want to engineer custom enzymes for biotechnology, medicine, or industrial chemistry.
Natural evolution is an incremental process. It works not by designing a protein from scratch but by modifying existing ones, often exploiting the flexibility of an ancestor that already had low-level activity toward several substrates. The case of NrnC and diDNase illustrates this principle with unusual clarity. The ancestral enzyme was evidently capable of giving rise to two different specialists, each tuned to a different kind of small nucleic acid molecule.
For modern protein engineers, the lesson is that simply mutating residues in a contemporary enzyme may not reproduce the path nature took. Sometimes the rules that govern an enzyme's evolution are buried deep in its history. To change a protein in a meaningful way, it may be necessary to think like an evolutionary paleontologist and consider the ancient forms from which contemporary proteins arose.
A Blueprint for Rewinding Evolution
The study also offers a methodological blueprint for other research groups. Ancestral sequence reconstruction is not new, but combining it with structural biology and laboratory resurrection is beginning to yield remarkable insights. The CSSB and DESY infrastructure allow researchers to bring together evolutionary analysis, protein production, and high-resolution structure determination under one intellectual umbrella.
The team's success also highlights the value of persistence. Initial mutation experiments failed, and only by stepping back and embracing a longer evolutionary view were the researchers able to make progress. In science, as in evolution, the path forward sometimes requires going backward.
Looking Ahead
More experiments will be needed to map the entire trajectory from ancestral enzyme to modern specialists. Now that they have the reconstructed protein in hand, the researchers can begin swapping in mutations that separate the two lineages and observing the consequences in real time. Each experiment will add a new detail to the picture of how this family evolved, and perhaps offer general principles that apply to many other proteins.
For nonspecialists, the work is a reminder that evolution is not only a force that shapes elephants and orchids, but also a sculptor of the molecular machines inside every living cell. By reviving a protein that had been lost to time, a team of scientists has shown that the past is not always beyond reach. Sometimes it can be reconstructed, studied, and made to reveal its secrets.
This article is based on reporting by Phys.org. Read the original article.
Originally published on phys.org
