Structural analysis of monomeric RNA-dependent polymerases revamped

Viruses are masters of disguise, constantly rewriting their own genetic instructions to evade detection. Among the vast, invisible world of RNA viruses, there is one protein that refuses to change its fundamental shape, even as its genetic code mutates wildly. This protein is a molecular machine called an RNA-dependent polymerase, the engine that copies the virus's genetic material so it can spread. Because the genetic code of these viruses changes so quickly, scientists have long struggled to trace their family history using DNA sequences alone; the signal of their ancient origins has been scrambled by time. However, the physical three-dimensional shape of a protein is far more stable than the letters of its code. Just as a stone tool retains its form long after the language of the people who made it has vanished, the architecture of this viral engine holds clues to a deep evolutionary past that sequences have forgotten.

A team of researchers at the Universidad Nacional Autónoma de Mexico has turned to these physical shapes to redraw the family tree of RNA viruses. Instead of comparing the letters of the genetic code, they compared the actual 3D structures of the polymerase engines from a vast array of viruses. They gathered data on over one hundred different viral engines, mixing structures that had been captured in laboratories with high-resolution models generated by powerful computer programs. By lining up these structures side by side and measuring how well they fit together, the team constructed a new map of viral relationships. This approach allowed them to see connections that were invisible to traditional methods, revealing how different groups of viruses are related and challenging some of the current rules used to classify them.

The study began by assembling a diverse collection of these viral engines. The researchers selected structures from viruses that infect everything from humans and plants to bacteria and fungi. Since experimental structures are difficult to obtain for many obscure viruses, they relied heavily on computer predictions, using advanced software to build 3D models of the polymerases based on their genetic sequences. They were careful to ensure these digital models were reliable, discarding any that looked uncertain or poorly formed. In total, they analyzed 107 structures, representing 103 different viral genera. This broad sample included viruses from nearly every major branch of the viral kingdom, providing a much wider view than previous studies, which had been limited mostly to viruses that cause human disease.

When the researchers compared these shapes, they found that the physical structure of the polymerase revealed a clear family tree. The resulting map showed distinct branches that largely matched the current official classification of viruses, grouping together those that share a common ancestor. For instance, the engines from viruses with positive-sense RNA genomes formed one major cluster, while those with negative-sense genomes formed another. However, the study also uncovered cracks in the current system. Two major groups of viruses, known as Lenarviricota and Duplornaviricota, did not form single, unified families in this new tree. Instead, their members were scattered across different branches, suggesting that the current way these groups are defined might not reflect their true evolutionary history. The researchers propose that these classifications need to be revised to better match the physical reality of the viral engines.

Beyond the big picture, the team discovered specific structural details that act as unique signatures for different viral families. They found that while the core of the polymerase engine is the same in all viruses, the parts of the machine that stick out before the main working area vary in predictable ways. Some viruses have a long, looping arm that wraps around the engine, while others have a bundle of helical springs that project downward. These subtle architectural differences appear consistently within specific groups of viruses, serving as a physical fingerprint that helps scientists identify where a new virus belongs on the family tree. These features were found even in viruses that had never been seen before, allowing the researchers to place newly discovered viruses into the correct taxonomic groups based solely on their predicted shape.

The study also addressed the challenge of viruses that do not fit neatly into existing categories. Some viruses have their genetic instructions arranged in a circular permutation, a rearrangement that flips the order of the engine's internal parts. Despite this unusual layout, the researchers found that the overall shape of these engines still grouped them with a specific family of viruses, suggesting they share a common origin with them. Similarly, they were able to incorporate structures from newly discovered viruses that infect fungi and bacteria, placing them into the broader evolutionary context. This ability to classify viruses based on their physical form, even when their genetic code is too different to compare, offers a powerful new tool for understanding the deep history of life on Earth.

While the new tree provides a clearer view of viral evolution, the authors caution that the deepest branches of the tree remain difficult to resolve with absolute certainty. The rapid mutation rate of these viruses means that the oldest evolutionary signals are faint, and even the most advanced structural comparisons struggle to pinpoint the exact order in which the major groups diverged billions of years ago. The study does not claim to have solved the mystery of the very first RNA virus, but it does provide a robust framework for understanding how the major groups we see today are related. By focusing on the unchanging physical shape of the viral engine rather than the rapidly changing genetic code, this research offers a stable foundation for exploring the ancient history of the virosphere, turning the invisible architecture of proteins into a readable map of life's past.