As a growing number of countries push for an independent investigation into the origin of the COVID-19 pandemic, many scientists around the world are already trying to uncover when, where and how the new coronavirus got into people.

Finding the source is important for preventing further reinfection, but scientists’ investigations — which include modelling, cell studies and animal experiments — are revealing how tricky pinpointing the source might be.

“It is quite possible we won’t find it. In fact, it would be exceptionally lucky if we land on something,” says Lucy van Dorp, a geneticist from University College London (UCL).

There is strong evidence that the virus originated in bats. The biggest mystery remains how it got from bats to people. Researchers overwhelmingly think that it’s a wild virus, which probably passed to people through an intermediate species. But no one has found the virus in the wild yet, so other explanations cannot be ruled out.

US President Donald Trump has fuelled suggestions that the virus might have leaked from a laboratory in Wuhan, where the outbreak started. There is no evidence for that claim.

Still, other world leaders have called for investigations into the outbreak's origin. The European Union and dozens of nations are supporting a draft proposal submitted to the World Health Assembly, the key decision making-body of the World Health Organisation, which is holding a virtual meeting with member states today and tomorrow. The proposal calls for “scientific and collaborative field missions” to “identify the zoonotic source of the virus and the route of introduction to the human population, including the possible role of intermediate hosts”.

The only way to say with confidence which animal the virus came from is to find it in that species in the wild, says Arinjay Banerjee, a coronavirus researcher at McMaster University in Hamilton, Ontario. “Other approaches will only give you anecdotal evidence,” he says.

But given that the virus has spread so widely among people, even detecting the virus in animals will not necessarily confirm their role as intermediate hosts as they might have been infected by people, says Li Xingguang, who studies viral evolution at Wuhan University of Bioengineering. “The situation is very complex now.”

Bat origin

Researchers’ first started looking at the virus’s genome to see whether they could match it to pathogens found in other animals. In late January, a few weeks after researchers sequenced the SARS-CoV-2 genome, scientists at the Wuhan Institute of Virology posted online the entire sequence of a coronavirus that had been stored in their lab since being discovered in intermediate horseshoe bats ( Rhinolophus affinis ) in Yunnan province in 2013. That genome, named RATG13, was 96% identical to SARS-CoV-2, making it the closest known relative and strongly suggesting the new virus originated in bats.

Computational biologist Francois Balloux and his team at UCL, including colleague van Dorp, and other teams are searching genomic databases of animals looking for coronaviruses that are an even closer match.

Although the 4% difference between the genomes of SARS-CoV-2 and RATG13 still represents some 50 years since they last shared a common ancestor, says van Dorp. The divergence is another piece of evidence that suggests that SARS-CoV-2 could have passed to people through an intermediate species.

Intermediate host

Pangolins were among the first animals suspected of being the intermediate. Two teams in China reported that they’d found similarities between SARS-CoV-2 and coronaviruses isolated from tissue of Malayan pangolins ( Manis javanica ) that had been confiscated. Trading pangolins is illegal in China.

The pangolin coronaviruses turned out to be too distant to be direct ancestors of SARS-CoV-2, but the fact that they are the only wild mammals besides bats known so far to be living with coronaviruses similar to SARS-CoV-2 suggests they can’t be ruled out as an intermediate source.

Scientists are looking for similar coronaviruses in other animals, too. The ancestor of SARS-CoV-2 could be lurking in tissue samples that are stored in a lab, says Aaron Irving, an infectious-diseases researcher at Duke-NUS Medical School in Singapore. “Many labs have samples sitting in their freezers,” he says.

Irving plans to collaborate with researchers at the Chinese Academy of Sciences (CAS) Xishuangbanna Tropical Botanical Garden in Yunnan to test tissue samples from wild mammals collected by wildlife surveillance programmes for coronaviruses that might be closely related to SARS-CoV-2. He is also about to start a new lab at the Zhejiang University-University of Edinburgh Institute in Haining, and plans to look for coronaviruses in bats, tree shrews, civets and other mammals, where permitted. But in February, China introduced a ban on wildlife farms and many are struggling to keep their civets alive, he says. “It may be too late when I get on the ground,” says Irving.

Genome clues

Examining the SARS-CoV-2 genome could also reveal clues about possible intermediate hosts. Over time, viruses often start encoding their proteins using similar patterns of nucleotides to their host’s, which helps the virus adapt to their new environment. Researchers at UCL are using machine learning to tease apart patterns in the genetic code of SARS-CoV-2 that could predict which animals it might have adapted to.

But other researchers urge caution about this approach. In the early days of the pandemic, scientists at Peking University Health Science Center noted similarities between the protein-coding patterns of SARS-CoV-2 with those preferred by two snake species. The theory that a snake could be an intermediate host was quickly refuted by other researchers who said that the small sample size and limited data meant that the observed patterns were probably down to chance.

Growing the virus in animal cells is one way to test whether the pathogen has adapted to a new host. Shi Yi, a microbiologist at the CAS Institute of Microbiology in Beijing, plans to introduce an inactivated version of RATG13, in various animals, such as bats, cats monkeys and pigs, and see whether the virus develops a similar pattern of mutations to SARS-CoV-2 over time. If similarities emerge, that could reveal which animals the virus adapted to before it jumped to people.

List of suspects

Determining which animals SARS-CoV-2 can infect is another way to narrow down the possible intermediate sources. “Knowledge on the susceptibility of different species and potential routes of transmission between animals could give us clues about the likely candidate host or intermediate host in China,” says Bart Haagmans, a virologist at Erasmus MC in Rotterdam.

Research so far suggests many species can be infected. In lab experiments, cats, fruit bats ( Rousettus aegyptiacus ), ferrets, rhesus macaques and hamsters have been shown to be susceptible to SARS-CoV-2. Outside the lab, animals including pet cats and dogs, tigers and lions at zoos, and farmed mink have also caught the virus — probably from people.

Researchers are also using computational models and cell biology to investigate animal susceptibility. SARS-CoV-2 typically enters cells through a receptor protein called ACE2. One unreviewed study, ¹ led by Christine Orengo, a bioinformatician at UCL, modelled the structure of ACE2 from more than 215 vertebrates and found that the receptor in many mammals, including sheep, chimpanzees and gorillas, engages well with the spike protein on the surface of the virus, which suggests that these animals might be susceptible to infection.

But modelling does not always correlate with the experimental evidence. For example, Orengo’s modelling suggests that horseshoe bats have a low risk of infection despite lab evidence that they can be infected. Another group, led by Yuen Kwok-yung, a microbiologist at the University of Hong Kong, has found ² that the virus replicates well in tiny organoids grown from intestinal stem cells of Chinese horseshoe bats ( R. sinicus ).

It’s useful to know which animals are susceptible, to manage the risk that they might become virus reservoirs and possible sources of infection in people, says Michelle Baker, a comparative immunologist at the Commonwealth Scientific and Industrial Research Organisation in Geelong, Australia. But when trying to narrow down the culprit, it seems sensible to focus on those animals in close contact with bats, she says.

Animals at wildlife farms in China are one of the first places to look, says Peter Daszak, president of the non-profit EcoHealth Alliance in New York City. Farms stock many captive-bred animals, from civets to raccoon dogs and coypu, a large rodent, often living close to livestock such as pigs, chickens and ducks. “These farms are usually wide open to bats, which feed at night above the pens, and some of which roost in the buildings. They are also usually linked to people’s houses so that whole families are potentially exposed,” says Daszak, who has visited many villages, wildlife markets, bat caves and farms in southern China over the past 15 years.

“The opportunities for these viruses to spill over across a very active wildlife–livestock–human interface is clear and obvious,” he says.