By Thabata Oliveira |Agência FAPESP* – A global mapping project revealed tens of thousands of previously unknown viruses circulating in environments with high human traffic, such as transit stations, hospitals, and sewage systems. The study, detailed in an article published in the journal Nature Communications, analyzed 2,922 samples collected in 102 cities across 31 countries, including São Paulo, Brazil. An international team, including Brazilian researchers, used this data to compile the UPVAtlas (Urban and Peri-urban RNA Virus Atlas), cataloging 54,945 viral species, 77% of which had never been described before. The study sheds light on vast, previously unknown biological diversity, revealing the complexity of microorganisms invisible in everyday city life.

The research focused on RNA viruses, which store their genetic instructions in RNA molecules rather than DNA. These viruses cause diseases such as dengue, Zika, the flu, yellow fever, and, most recently, COVID-19. Much of what is known about this group comes from studying these diseases.

“We were familiar with several disease-causing viruses, but we greatly underestimated viral diversity in nature. When you search for a virus based on a disease, there’s already a bias because all viruses that don’t cause disease end up being ignored. Once we’re able to analyze what’s present in a sample without that bias, we gain a broader view of viral diversity,” says Antônio Pedro Camargo, a professor in the Department of Biochemistry at the Institute of Chemistry of the University of São Paulo (IQ-USP) and one of the authors of the study. He is also a member of the Biology of Bacteria and Bacteriophages Research Center (RIDC B3), a Research, Innovation, and Dissemination Center supported by FAPESP.

Metatranscriptome

The MetaSUB consortium, which investigates the microbiome of urban environments, collected most of the material – 1,862 of the 2,922 samples. Between March and July of 2020, during the pandemic, the consortium collected the samples using a standardized protocol. A swab was rubbed against frequently touched surfaces, such as handrails, benches, and ticket machines. The remaining samples were retrieved from a public sequencing database and represent various aspects of the urban environment, including soil, sediment, freshwater, sewage, and greenhouses.

The materials were analyzed using metatranscriptomics, a technique that allows for the simultaneous sequencing of all RNA molecules present in a sample. “Imagine you pick up a handful of dirt. In there, you’ll find bacteria, bacterial viruses [bacteriophages], other microorganisms, insects – everything you can imagine,” Camargo explains. Rather than isolating each organism, the researchers extract and sequence the entire set of RNA from the sample.

The result is a large volume of unidentified fragments. “You’ll have a bunch of genome fragments, and at first, you won’t know what they are. They’re just sequences of letters. The challenge is figuring out what each piece of the genome corresponds to. Is it a bacterium? Is it a virus? If it’s a virus, what type?” the researcher explains.

This screening was performed using geNomad, a bioinformatics tool developed by Camargo. According to him, the program “allows us to process millions of genetic material sequence fragments and determine which correspond to viral genomes.”

To organize their findings, the researchers compared the sequences of an enzyme present in all RNA viruses, RdRP (RNA-dependent RNA polymerase), which is responsible for viral replication, and reconstructed the phylogenetic relationships among them. Some sequences were so different from anything previously described that the authors proposed creating new categories to accommodate them, including two phyla, a new class, and several unclassified groups. The proposal is preliminary. “This is all computational. You need more information to formally state that these are, in fact, completely new groups,” Camargo notes.

However, the survey revealed only a fraction of the actual microbial diversity. The accumulation curves, which measure the number of previously unknown viruses that appear with each new sample analyzed, did not stabilize in virtually any type of sample. This is a sign that the nearly 55,000 cataloged viruses represent only a fraction of what exists in these locations.

“There’s a very high level of viral diversity – greater than we previously knew. I think that’s the main point of the article,” says Camargo.

Hosts

This discovery does not mean that cities are teeming with new viruses capable of causing human disease. Only about one-third of the sequences had potential hosts identified, and more than half of those corresponded to viruses that infect bacteria.

However, a small portion has the potential to infect vertebrates. By cross-referencing the data with the Global Virome in One Network (VIRION) database, which compiles documented associations between viruses and vertebrates, the authors identified 60 viruses with a high probability of infecting humans and 47 that could infect other mammals. Of the 60 viruses, 19 were found in urban environments, and 34 were found in animal samples.

The study also identified 166 bacteriophages, which are viruses that specialize in attacking and destroying bacteria, with a history of infecting four strains from the ESKAPE group, which includes clinically significant microorganisms such as Staphylococcus aureus. According to the study, this finding could lead to research on using these viruses to combat these pathogens through a treatment known as phage therapy.

Surveillance

In general, urban surfaces, such as streets, banks, stations, and stores, exhibited lower viral diversity than natural environments, such as soil and sediment. According to the authors, this indicates that human activity tends to homogenize these communities. However, sewage was an exception, exhibiting one of the highest levels of diversity in the survey. This finding reinforces the potential of sewage for environmental monitoring.

Among hospital samples, São Paulo, Brazil, and Islamabad, Pakistan, had a higher abundance of vertebrate-pathogenic viruses than Baltimore, United States, and Seoul, South Korea. Camargo urges caution when interpreting these data: “It’s very difficult to infer any biological explanation from that. It may simply be a sampling issue.” The study itself acknowledges this limitation; some countries are represented by only one environment, making it impossible to separate the effect of location from that of the country.

UPVAtlas compiles sequences from different cities and environments into a single catalog, creating a starting point for future comparisons. It will be possible to compare a city’s current viral community with its future one and investigate how viral communities change depending on the environment and location.

Understanding this diversity is important to Camargo precisely because viruses are constantly evolving. “This process is closely related to their ability to infect organisms because viruses only survive if they can reproduce within host cells,” he says.

The article “Diversity and distinctive characteristics of the global RNA virome in urban and peri-urban environments”can be read at https://nature.com/articles/s41467-026-73605-z .

* Thabata Oliveira (https://bv.fapesp.br/pt/pesquisador/745574/thabata-fernanda-oliveira) is a FAPESP Scientific Journalism fellowship holder (https://bv.fapesp.br/pt/bolsas/233629 ) affiliated with IQ-USP