There seems to be no pattern to where humans pick up new viruses

A virus that normally infects animals makes the jump to humans, whose immune systems have never seen it before. It suddenly sweeps across the globe, leaving death and chaos in its wake. We’re living with that reality now and have gone through it previously with HIV, SARS, MERS, Ebola, Hanta, and various flu viruses that have threatened humanity in just the past few decades.

While there are many organizations that try to stay on top of threats of emerging diseases, it would be helpful if we could identify major sources of potential threats. If, for example, we knew that certain species were more prone to carrying viruses that could make the jump to humans, we could potentially survey the viruses found in those species, identify major threats, and potentially even develop therapies or vaccines in advance.

But a study published recently in PNAS suggests there’s no real pattern to where humans are picking up new viruses. Instead, groups with lots of species tend to have lots of viral species, and those make the jump to humans largely in proportion to the number of species.

Zoonotic risk

A disease that can be transmitted from animals to people is technically called “zoonotic.” While there are a variety of diseases that incorporate time in another species as part of their lifestyle—malaria is a classic example—the risk we’re concerned about is a virus that normally circulates within a non-human species but evolves the ability to spread within humans and leaves its original host behind.

These sorts of events are relatively common. Flu viruses seem to hop among us and our agricultural species with some regularity. Other viruses, like members of the hantavirus family, seem to frequently make the jump to humans without ever establishing the ability to spread from human to human.

It’s the latter feature that creates the risk of a global pandemic. Two earlier coronaviruses, SARS-CoV-1 and MERS, didn’t spread among humans as effectively as SARS-CoV-2, allowing containment methods to halt their spread before a pandemic could develop.

Are there any species that might be especially good launch pads for a pandemic? A couple of hypotheses suggests this could be the case. One hypothesis is that evolutionary distance matters. A virus that normally circulates in a species that’s related to humans is more likely to have components that can interact more effectively with the proteins that are present in human cells. If this were the case, we’d probably expect to see more zoonotic jumps taking place from viruses that infect our fellow primates.

An alternate idea has come out of the fact that this doesn’t seem to be consistently true. Bats, for example, have “gifted” humans with such distantly related viruses as SARS-CoV-1 and Ebola, and they’re not especially closely related to us. As a result, researchers have hypothesized that there might be what have been called “special reservoirs,” or species that, for ecological or lifestyle reasons, have ended up with viruses that can adapt more readily to human hosts. These special reservoirs could simply be more likely to live in close proximity to humans, raising the risks of transmission.

Two Glasgow researchers, Nardus Mollentze and Daniel Streicker, decided to conduct a test of these two hypotheses by figuring out whether there were any groups of species that were more likely to spread viruses to humans.

Building trees

To do so, Mollentze and Streicker built a comprehensive database of every virus that has been reported to make the jump to humans, as well as the host from which it jumped. In all, there were 415 different viruses that had a host assigned and could be used for the analysis (that’s out of 673 known virus species). These were spread across 30 families (the designation two levels above species) and had made their way out of 11 different orders of host species (an order is the level above family).

On their own, the results would seem to point to the special reservoir model, as hoofed ungulates (like our agricultural animals) and rodents collectively accounted for half the viruses that had transitioned to human hosts. But things got more complex when the authors tried to analyze the properties of a virus that made it more likely to make this transition. The best combination of properties, which could explain about half the probability of a zoonotic jump, was dominated by things like transmission through insects and a relatively simple replication cycle inside cells.

An  while the host’s order on the evolutionary tree appeared to matter at first, it mattered much less once the authors adjusted for a critical factor: how many individual species make up that order. For example, rodent and ungulate species may transmit more viruses to us, but there are a lot of species in these groups. If you adjust the rate by species number, the effect largely goes away. If you also control for the fact that we’ve identified far more virus species in mammals than in birds, then the effect becomes little more than statistical noise. The probability that a group of species will transmit a virus to humans becomes a function of how many species are in that group.

This is inconsistent with the special-reservoir hypothesis. But things don’t look great for the evolutionary explanation, either. While the zoonotic risk dropped as you got further from primates, this accounted for less than 1 percent of the overall risk.

In fact, if you simply estimated the number of zoonotic jumps based on the species number, groups that seemed threatening start to look fairly mundane. Rodents, for example, would be expected to have given 42 viruses to humans; we’re aware of 41 instances where that took place. Bats would be expected to have transferred 28 viruses to us but have only sent 22 of them. The one exception is, again, the ungulates, which seem to send viruses our way at rates above what we’d expect.

Now what?

The hope was that, by identifying the rules of zoonotic transfers, we could identify groups of species that have an elevated risk of causing problems and thus could be subjected to more careful monitoring. This analysis suggests that these groups might not exist. It doesn’t rule out the possibility that there are groups of species below the order level that are hotspots for zoonotic transfers. But at this point, the number of viruses transferred per group is likely to be small and might not stand out from statistical noise.

That said, some species/virus combinations are notable. For example, while bats are notable for having been the source of SARS-CoV-1 and Ebola, they’re actually most likely to transfer a new species of rabies virus to humans. Other primates are a major source of adenovirus and Dengue species, while rodents tend to transfer hantaviruses and arenaviruses.

While this isn’t especially good news for targeted surveillance efforts, that might not be bad news overall. Having obvious targets might mean we over-focus on those, leaving us vulnerable to risks we hadn’t anticipated.