The microbiology toolbox includes techniques to induce mutations in viruses that give the microbes new powers. Scientists perform these manipulations for many reasons, including wanting to understand how the microbes evade detection by our immune systems. But adding capability to a pathogen carries obvious risks, especially if this “gain of function” involves enhanced virulence or infectiousness. Escape from a lab, by accident or design, is a possibility. So why do it? Some researchers argue the work can offer a peek at what a virus can do before it goes into the natural world and poses a threat to people.

Controversy over gain-of-function research has generated academic papers, conferences and even a moratorium in 2014, when the U.S. government paused funding for three years until steps could be taken to ensure the safety of the procedure. Debate about gain-of-function experiments continues in the latter phases of the pandemic as thoughts turn to the “next one” or a possible second act for COVID-19. Science policy makers must wrestle with defining the rare instances in which the benefits of experiments that enhance a virus’s capacity to survive and flourish in human hosts outweigh any risks.

Densely technical discussions often bog down over the very definition of gain of function. Recently, semantics were front and center in the debate over whether National Institutes of Health–funded work at the Wuhan Institute of Virology (WIV) in China constituted gain-of-function research, a contention denied by the U.S. agency. The WIV has also been the focus of a revived dispute over whether SARS-CoV-2, the virus that causes COVID-19, escaped from its facility.

Here are a few basic answers to questions about why an obscure technical term now receives so much attention.

What is gain of function research?

Techniques to enhance some aspect of an organism’s functioning are commonplace in research and applied to everything from mice to measles. One typical application of this approach is tweaking mouse genes to generate more of a protein that limits fat deposition.

But that is not the kind of gain-of-function study that raises fears among scientists and regulators. The high-risk practices are those that create mutations to examine whether a pathogen becomes more contagious or lethal as a means of estimating future threats.

Some experts acknowledge the critical differences between the two types of studies. One proposed term to represent the more threatening subset of this research is “potential pandemic pathogens,” says Marc Lipsitch, a professor of epidemiology at the Harvard T. H. Chan School of Public Health. That phrase “singles out the name and reason for being concerned,” he adds. It has not caught on in common usage, however, returning only about 8,500 results in a Google search, compared with 13.4 million for “gain of function.”

Making this distinction is important for a few reasons, Lipsitch says. When the U.S. government placed the 2014 moratorium on “gain of function research,” some of the studies that were affected carried no obvious risk of setting off a pandemic.

What is the purpose of this research?

Knowing what makes a microbe more dangerous enables preparation of countermeasures, says Lipsitch, who is one of 18 signatories to a May 14 letter, published in Science, that calls for the investigation of a SARS-CoV-2 lab spillover as one of several possible explanations for the origins of the COVID-19 pandemic. He points to the difficulties of studying viruses for the development of vaccines and treatments without doing experiments in a mouse or in other nonhuman animals. There is, Lipsitch says, a “direct path from doing that research to gaining public health benefits,” enabling a balancing of risks and potential benefits.

The riskier version of gain-of-function research creates viruses with abilities they do not have in nature. In two separate studies in 2011, scientists famously and controversially did just that with the H5N1 influenza virus, or “bird flu,” resulting in a version capable of airborne transmission among ferrets. The naturally occurring virus does not have this ability. Making mammal-to-mammal transmission easier set off alarm bells and triggered discussion of a U.S. moratorium.

In 2015 researchers engineered a hybrid pathogen that combined features of the original SARS virus (SARS-CoV) that infected humans in the early 2000s with that of a bat coronavirus. Most bat coronaviruses cannot infect the cells lining the human respiratory tract. This experiment was intended to mimic what would happen if a third species served as a mixing vat for the bat and human viruses to exchange genetic material. The result was a pathogen that could enter human cells and also cause disease in mice. Reactions to this work were polarized, as demonstrated by experts quoted in a 2015 article in Nature: one said that all the research did was create a “new, non-natural risk” among the multitude that already exist, while another contended that it showed the potential for this bat virus to become a “clear and present danger.”

Experts in the latter camp argue that gain-of-function virus studies can presage what will eventually happen in nature. Speeding things up in the lab gives researchers firsthand evidence about how a virus might evolve. Such insights  could drive predictions about future viral behaviors in order to stay a step ahead of these pathogens.

That calculation must be made on a case-by-case basis, Lipsitch says. “There is not one-answer-fits-all,” he adds. But the key question to address in this complex computation is “Is this work so valuable for public health that it outshines the risk to public health in doing it?”

Lipsitch was “very outspoken,” as he puts it, about the influenza-ferret study, and he led the effort for the 2014 moratorium on similar gain-of-function work. “I did that because I thought that we need to have a real accounting of the benefits and risks,” he says. “I had a view that the benefits were very small, and I still have that view.”

The moratorium was lifted in 2017.  A U.S. government review panel later approved a resumption of funding for more lab studies involving gain-of-function modifications of bird flu viruses in ferrets. Conditions of the approvals, according to reports, included enhanced safety measures and reporting requirements.

As for SARS-CoV-2, the virus of most urgent interest right now, the NIH released a statement on May 19 that neither the agency nor its National Institute of Allergy and Infectious Diseases has “ever approved any grant that would have supported ‘gain-of-function’ research on coronaviruses that would have increased their transmissibility or lethality for humans.”

What are the risks?

Predictions based on gain-of-function studies may be hypothetical, but lab breaches in the U.S. are not. Serious violations are uncommon and have almost never resulted in a pathogen being released into the community. But 2014 showed why human error may prove to be the biggest wild card in planning these experiments.

Several lab accidents that year endangered researchers and set off waves of uneasiness. These incidents were not gain-of-function mishaps, but they demonstrated the potential threats posed by a biosafety lab—whether from negligence or malfeasance. In 2014 about 75 Atlanta-based employees at the U.S. Centers for Disease Control and Prevention learned about their potential exposure to anthrax after safety practices were ignored. Also, several long-forgotten vials of freeze-dried smallpox—a pathogen long thought to be stored in only two places, one in Russia and one in the U.S.—turned up during a cold-storage cleanup at the NIH that year. And the CDC made news again a month later, after it sent out vials of a relatively benign influenza virus contaminated with the much more deadly H5N1 avian flu virus. The possible reason, as reported in Science, was that a researcher was “overworked and rushing to make a lab meeting.”

Michael Imperiale, a professor of microbiology and immunology and associate vice president for research and compliance at the University of Michigan, co-authored a 2020 editorial about gain-of-function studies that said that the key to planning them is to have proper mechanisms to ward off the threats of accidental or intentional harm. “If proper biosafety procedures are in place and proper containment is used, the risks can be mitigated substantially,” he says. Biosafety level 4 (BSL-4) labs have the highest containment precautions in place, and the U.S. currently has 13 or more such facilities planned or in operation. Research on the novel coronavirus is handled in labs one notch down: BSL-3.

In their editorial, Imperiale and his co-author Arturo Casadevall, editor in chief of mBIO, wrote that even predicting the threat level of an accidental release is difficult. After publication of the studies of ferret-to-ferret transmission of engineered H5N1, two groups tried to predict what would have happened if this virus had escaped into the human population. One team, Imperiale and Casadevall wrote, predicted an “extremely high level” of transmission. The other, from one of the labs involved in the ferret-influenza work, concluded otherwise.

In the context of the COVID-19 pandemic, the authors of the editorial wrote, the source of a pathogen—whether from nature or a lab—does not change how the world should prepare to respond to it. But gain-of-function experiments should be governed by transparency in planning the research, a “rededication” to biosafety and a strong surveillance program to capture breaches.

What alternative techniques are available to test a potential viral threat?

If a virus has already moved from an animal host to humans, gain-of-function research may be unnecessary, Imperiale says. “In these cases, there may be animal models that serve as useful surrogates for humans” in testing the virus’s effects, he says.

Researchers can also  test the capacity of virus proteins to engage with different kinds of cells. Software can predict how these proteins might interact with various cell types or how their genetic sequences could be associated with specific virus features. Also, if the researchers use cells in a lab dish, the viruses might be designed not to replicate.

Another option is loss-of-function research. Using versions of a virus with less pathogenic potential is another way to unlock that microbe’s secrets. Still, highly pathogenic forms can be quite different from their less threatening counterparts—for example, they may differ in how often they replicate—possibly limiting the usefulness of such studies.