Researchers at the University of Miami have identified a group of previously unknown viruses that bridge an evolutionary gap between viruses and cellular protein compartments, a discovery with important implications for both biotechnology and scientists’ understanding of the origins of life.
The findings, published in Nature Communications , challenge the theory that these cellular compartments, called encapsulins, simply evolved from viruses. Instead, the researchers found evidence that viruses and encapsulins evolved from one another repeatedly over millions of years.
The researchers focused on a group of proteins that adopt a specific three-dimensional structure called the HK97-fold and assemble into protein shells. They are found in both viruses and encapsulins. The newly identified viruses—whose genomes are smaller and simpler than those of other viruses that form HK97-fold protein shells—share features with both viruses and encapsulins, making them a possible missing link between the two that could shed light on how complex biological systems first emerged.
“What makes the HK97-fold proteins so interesting is that they can form shells that enclose biochemical reactions in cellular lineages, archaea, and bacteria. But they also serve genetic purposes in viruses, storing and protecting their viral genome,” said Abelardo Aguilar , the first author and a Ph.D. student in the Department of Biology at the University of Miami College of Arts and Sciences . “These proteins build compartments with two different roles, which connects directly with the study of the origin of life and its early evolution because if you think about it, life at its core is about enclosing reactions and also enclosing information that can be inherited so those reactions can happen in the future.”
To identify these new viruses, Aguilar worked under the supervision of Antoni Luque Santolaria , an associate professor in the Department of Biology who has secondary appointments in the Department of Physics in the College of Arts and Sciences and the Department of Chemical, Environmental and Materials Engineering in the College of Engineering , and the senior author of the study, which included collaborators across different institutions. The team used bioinformatics tools, biophysical models, and AI-based tools, including protein language models and AlphaFold, to mine structural information from databases containing billions of DNA sequences. The sequences in the databases—which included IMG/M , a database affiliated with the U.S. Department of Energy’s Joint Genome Institute and the University of California, Berkeley—come from samples of soil, seawater, wastewater, the human gut, and other environments that were collected around the world.
“The majority of viruses cannot be grown in the lab. But now through sequencing, we have access to information on viruses for which we don’t know the host, that we can’t culture in the lab, but that are everywhere,” Luque explained. “That means there are multiple questions in relation to evolution and how viruses have adapted to the world that it’s now possible to explore bioinformatically.”
The Nature Communications article also outlines the potential molecular mechanism by which viruses and encapsulins may have evolved from one another. This mechanism shows how procapsids, which are immature viral protein shells (capsids), could allow evolutionary transitions and carry both genomic and metabolic roles.
“One implication of this research is the possibility that viral capsids could also facilitate biochemical reactions, which would open up the potential roles that capsids have in environments as a whole,” Aguilar explained.
These findings raise new questions about viruses that Aguilar and Luque said warrant further investigation.
Antoni Luque Santolaria Photo: Jenny Abreu/University of Miami
“When we’re looking at viruses, we tend to look only at the particles that are infective, and this study points out that we might need to start putting more emphasis on the non-infective versions of the virus and what other functions those viral-like particles may have that we have overlooked in the past.”
The research, which was supported by funding from the Gordon and Betty Moore Foundation and the U.S. National Science Foundation , demonstrates the potential of new AI-based bioinformatics tools to explore other unanswered questions about viruses.
“ If you have the right question and you pair it with the right tools and physical insights, it gives researchers an opportunity to look at viruses in a way that we didn’t have an opportunity to do before,” Luque said. “There are evolution questions, but there are also questions about emerging pathogens or emerging viruses. The set of tools that we use can be moved forward in different contexts.”
Nature Communications
Data/statistical analysis
Small viruses reveal bidirectional evolution between HK97-fold viruses and encapsulins via procapsids
14-Oct-2026