Gain of Fiction Research of Concern
The mortality rate is the only thing that will be improved by banning essential virology and vaccine research
In May, I was in Montréal giving a talk at McGill University. I was invited to give the keynote address at the Department of Microbiology and Immunology’s Graduate Research Day. It is a huge honor to be invited by students to give a talk, since it is a way for middle-aged professors to flex that we are still relevant to young, cool people. Because I also think it is really important that my Canadian colleagues, particularly those who are just beginning their careers, understand how they will be impacted by the destruction of science occurring in the US, I now devote a substantial portion of my talks to spelling out exactly how the would-be authoritarians in charge are combining disinformation, draconian funding cuts, and abysmal policy decisions to destroy science as part of the larger efforts to annihilate American democracy.
My talk at McGill was aptly timed, because while I was standing at the podium speaking, I saw my phone light up out of the corner of my eye. Then a few minutes later, it lit up again. And again and again. As it turns out, a new policy decision dropped while I was talking: President Trump’s Executive Order, “Improving the Safety and Security of Biological Research”. Journalists and colleagues alike wanted to know, what did I think of it?
Not highly. The EO is written so vaguely that, if interpreted broadly, it would shut down virtually all biomedical research. As usual, it’s also justified with more “Gold Standard” bullshit about restoring public trust, because nothing builds confidence in science like an ideologically-motivated royal decree that cancels it, endangering millions and wasting billions of dollars on terminated projects in the process.
However, nobody knew how it would be interpreted or enforced. NIH was not able to give guidance, probably because the program officers who haven’t been fired weren’t given any guidance themselves on how to advise applicants and grantees. Until a few weeks ago, when NIH Director Jay Bhattacharya’s office dropped this notice.
This policy provides only slightly more clarity. Not in terms of the how the banned “dangerous gain-of-function research” is defined—which remains choose-your-own-adventure for the regulators at NIH who will oversee its enforcement—but regarding what those regulators will do. Specifically, they are going to terminate funding and in-kind support for projects conducted at foreign institutions that are deemed to violate the policy. They are also going to suspend all NIH funding and support for projects that are designated as forbidden by whoever is charged with interpreting the Executive Order.
Now this is starting to make more sense. The Trump administration has been producing some extremely wild and not remotely true Lab Leak content to justify the evisceration of American infectious disease research. The story goes that Tony Fauci funded “dangerous gain-of-function” research at the Wuhan Institute of Virology to create SARS-CoV-2, this leaked from the lab, and we got the COVID pandemic. Therefore, the NIH needs to defund virology and vaccine research. Also, all of us virologists are irresponsible gain of function junkies whose career successes are defined by the egregious riskiness and pointlessness of our experimental work, so our “bioweapon agent” factories need to be shut down. These lies have already been used to shutter critical emerging infectious diseases research programs. My colleague, biodefense and health security expert Gigi Gronvall, even gave this campaign a name: Gain of Fiction. It’s an apt term, because a false narrative will be the only thing the US will gain with this policy. The losses, on the other hand, will be profound.
What is gain-of-function research of concern?
This is a hard question to answer because it has not been well-defined. Technically, gain-of-function research is any experiment where you confer a “function” or ability on an organism that it didn’t have before. If you stick a gene for a protein that makes a cell grow fluorescent green into yeast, your now-fluorescent green yeast have gained green fluorescence function. Even though this is a gain of function, it’s not one that presents a danger to anyone.
Gain-of-function research of concern (GOFROC) is what government regulation can and should address. GOFROC is a type of microbiology research that results in engineering or selecting for a pathogen that is more transmissible, virulent, immune evasive, or countermeasure resistant. It is a type of dual use research of concern (DURC), which is anything that can have both beneficial and harmful or malicious applications. Making a bacterial pathogen resistant to last-resort antibiotics, making a virus capable of being transmitted by a different route, or expressing or enhancing a dangerous microbial toxin are all examples of GOFROC. These experiments can and do yield benefits. In fact, the benefits of GOFROC have been immense in terms of both developing countermeasures for infectious diseases and advancing science in general, but they also present significant risks. A “lab leak” of a more antibiotic-resistant strain of Strep or a more transmissible flu virus could be catastrophic.
Although that seems straightforward, it’s actually very, very complicated. No two experiments or pathogens or experimental systems are the same, so risk assessments are often extremely technically nuanced. Often there are trade-offs, as well: making a virus more transmissible may make it less pathogenic, so how do you assess the overall change in risk? Who has the responsibility and expertise needed to make these assessments in a fair and objective way? And how do you define GOFROC clearly enough to carry out practical and effective oversight? Our prior approaches to regulating GOFROC and DURC did not successfully resolve these questions, thus opening the door for this broad Executive Order and NIH policy.
A brief history of GOFROC
In 2005, a team led by now Acting NIAID Director Jeffrey Taubenberger resurrected the 1918 pandemic flu virus using a technique called reverse genetics. Flu reverse genetic systems work by cloning the different flu segments into circular pieces of DNA called plasmids. You stick all these plasmids into a cell, viral proteins get made and genome segments get replicated, and these get assembled into infectious flu viral particles that you can then use for your virology experiments.
Opponents of virology research contended from the outset that there were no benefits from this research, but I strongly disagree. By reconstructing 1918, we had the unprecedented ability to study the virus that caused a generational pandemic. We learned a lot about how 1918 causes disease and how we respond to and develop immunity to it. It was relevant at the time, as all of the H1N1 influenza that circulated in the human population since 1918 was a descendant of the OG pandemic strain until 2009, when they were replaced by the 2009 pandemic strain. Taubenberger has used 1918 to test vaccine candidates, including the inactivated vaccine that he has been given $500M in funding to develop. It is also very useful as a comparator for assessing risk from H5N1 and other avian subtypes with pandemic potential, since 1918 itself was an avian virus that had recently acquired the ability to spread in mammalian hosts when it emerged. As human-to-human spread is the barrier to H5N1 becoming a pandemic virus, we can learn a lot from studying how 1918 overcame this obstacle naturally as it adapted to human hosts. However, because it is a dangerous pandemic virus, 1918 is heavily regulated. It is a Select Agent, can only be used by personnel with security clearances in approved, frequently inspected, high security labs operating with enhanced biosafety level 3 (BSL-3) containment, and it is a federal crime to handle it without authorization.
At the time, there was a lot of debate about the ethical and safety risks of bringing back a pandemic virus that was no longer in this world (even though that wasn’t really true, since forms of it were still circulating as seasonal H1N1). That debate set the stage for what happened a few years later. In 2011, two flu researchers—Ron Fouchier at Erasmus Medical Center in the Netherlands and Yoshi Kawaoka at the University of Wisconsin—used different approaches to adapt H5N1 to spread by aerosols between ferrets. Fouchier presented his data at a conference and all hell broke loose. Some people claimed the studies should never have been done in the first place because of the potential for a flu pandemic. Some people thought they should not be published, due to the risk that someone would try to duplicate their methods to create a bioweapon. Ultimately, none of those things happened. The papers were published in full in Science and Nature, Fouchier, Kawaoka, and others have continued to study these viruses safely in appropriate biocontainment. These viruses have never leaked from a lab.
Some people have also suggested that these experiments were also more pointless virological recklessness, but that’s simply not true. We learned a lot from these studies, much of which is directly applicable to the current H5N1 situation we are facing, even though the H5N1 viruses Fouchier and Kawaoka worked with are very different from the ones circulating now. We learned that it actually doesn’t take much for H5N1 viruses to adapt to respiratory transmission in mammals. We also learned what mutations the virus might need to acquire to adapt to efficient replication and transmission by the respiratory route. This is particularly relevant, since one of them, PB2 E627K, has emerged independently in human H5N1 cases associated with this outbreak, including in the patient from Louisiana who died. These studies help us monitor natural respiratory adaptation that might occur in the H5N1 viruses that are circulating now in cattle and other animals and may give us a warning signal about emerging variants capable of spreading in humans. And we also learned that respiratory adaptation attenuated these viruses compared to the wild-type.
Nonetheless, virologists announced a voluntarily 1-year moratorium on GOFROC studies, defined as those that would enhance a respiratory pathogen like flu, SARS-CoV, or MERS-CoV to be more transmissible and/or pathogenic. In 2014, the White House stepped in and imposed a three-year moratorium on funding new grants that included GOFROC on respiratory pathogens. The moratorium was lifted when the White House Office of Science and Technology Policy (OSTP) unveiled the Potential Pandemic Pathogen Care and Oversight (P3CO) framework, which would govern the use of “PPP” (potential pandemic pathogens) and “ePPP” (enhanced potential pandemic pathogens). The P3CO defined GOFROC with ePPP as experiments that would be “reasonably anticipated” to increase transmissibility and/or pathogenicity in humans. Once again, the lack of a coherent definition hindered this policy’s effectiveness. It turns out that “reasonable anticipation” actually means a lot of different things to different people. Furthermore, the P3CO review process was not transparent, not implemented consistently, and had the effect of hindering essential research. The prospect of making a misstep, being erroneously accused of doing unsanctioned or reckless work, or having one’s funding paused or suspended for an opaque and indefinite review process, discouraged many researchers from pursuing critical work on flu and other pandemic viruses.
In 2023, the Biden OSTP put out a call for public comment about revising the P3CO to address some of these issues. I wrote a position paper in the Journal of Virology with about 80 of my colleagues calling for this process to be technically informed. In 2024, the new policy came out. It improved on a lot of things from the P3CO and did call for technically informed risk assessments. However, it was not to be. The new policy was cancelled by the Executive Order. And the Executive Order will be used to ban broad categories of research that the current administration finds undesirable. Everything is is GOFROC if it’s not MAHA.
GOFROC-eting to infinity and beyond
Now, I should probably disclose that the folks over at NIH who are working out how to implement the Executive Order as policy are not my biggest fans. Here’s NIH Director Jay Bhattacharya with his newly minted biosafety minion Bryce Nickels suggesting that I’m an anti-science bully who is jonesing to get my GOFROC on, because I objected to virology research being defunded for political reasons. Nickels co-founded an anti-virology lobbying group called Biosafety Now! and promptly recruited Bhattacharya as a board member. Together they continued a lengthy, biosafety- and lab leak-themed campaign of harassment against me and my co-authors, partly in retaliation for complaining about Nickels and his Biosafety Now! co-founder Richard Ebright defaming us for publishing peer-reviewed papers that indicated the COVID-19 pandemic began through zoonotic spillover at the Huanan Seafood Market and not a lab leak.
It’s pretty surreal to be accused of reckless virology by the NIH Director. It also feels kind of awkward, since it reveals that Bhattacharya doesn’t actually know that much about biosafety or virology research. I have explained many times that I study how the host responds to virus infection, so I don’t have any scientific reason to tinker with the viruses that I study. And while I do work with dangerous viruses in a high containment lab, they are naturally dangerous and I don’t modify them. I want to see how the host responds to the virus as it is in nature, so that we might understand how to treat the disease these viruses cause or improve vaccine effectiveness. I do not do GOFROC, nor do I have any plans to do GOFROC. And of all the things I care about in the laboratory, nothing is more important in virology research than biosafety. I have no intention of getting myself or any of my colleagues infected with any of the viruses I study and every move we make in the lab is designed to mitigate exposure risk while maximizing safety and security. Every virologist I know is very pro-biosafety, because that is essential to our work and our professional culture. Bhattacharya’s spurious accusations on the basis of Nickels’ ongoing unprovoked character review don’t suggest that these men are taking a very serious, informed, or evidence-based approach to implementing effective biosafety and biosecurity policy.
Unfortunately, it’s difficult to have reasonable discussions about the risks and benefits of different governance frameworks with people who openly despise you and lie about you. That’s a shame, because one would hope that Bhattacharya would have some of the humility he frequently feigns and make an effort to understand the research that the Executive Order might prohibit. It extends far beyond any previous definition of GOFROC. An incompetent or overzealous approach to implementing this policy would effectively crush entire scientific fields, although I’m pretty sure that’s exactly what Bhattacharya has in mind for virology. That’s what this policy will allow and I’ll explain why.
(a) Enhancing the harmful consequences of the agent or toxin
This seems pretty clear: no making agents or toxins more harmful. The problem is, what is “harmful”? There are different types of harm. There are also different degrees of harm. Let’s say I am working with the Cantell strain of Sendai virus. This is a paramyxovirus we use in the lab to induce interferon responses because it does that very well, it’s not very cytopathic, and it is heavily lab-adapted and isn’t thought to infect people or animals very efficiently. Sendai Cantell is adapted to grow well in eggs, so if you wanted to grow it in cells, you might need to pass it through your cell line of choice a few times to adapt it. So let’s say I do that and it becomes cytopathic (starts killing cells) in my Vero cells or whatever. That’s enhancing a “harmful consequence” to the cells I’m growing it in, but not to any actual person or animal. Would increasing cytopathic effect produced by a harmless virus in culture be considered sufficiently harmful that it would fall into this category?
Interpretation: almost all pathogen research can have “harmful consequences,” depending on how that is defined.
Possible effect: almost all pathogen research can be prohibited on a strategically focused “harmful consequence” argument
(b) Disrupting beneficial immunological response or the effectiveness of an immunization against the agent or toxin
This refers to making a pathogen or toxin more immune-evasive, although given this administration’s hostility toward vaccines, I have to wonder how much they really worry about this. Reducing a vaccine’s effectiveness is a pretty clear prohibition, but what the hell does it mean to “disrupt beneficial immunological response”? Again, that could mean a lot of things.
If its meaning is extended to any immunological response and not just cellular responses (B and T cells), it could mean that we’re not allowed to make viruses that will modify the host’s antiviral response. Viruses have evolved an impressive variety of ways to do this. Understanding the effects of this is a significant component of my own lab’s research and there are many ways to study it. One way to do this very safely is to use a reporter system to look at how viruses impact a key antiviral molecule called interferon. In my lab, we use the aforementioned Sendai Cantell virus to induce interferon, then observe how different proteins from Ebola are at suppressing it. Because we aren’t using infectious Ebola virus, we can study how host cells respond to individual viral proteins safely in BSL-2 containment. This does nothing to make Sendai more dangerous. There is no way this experiment would harm our ability to mount an antiviral response against Sendai or any other virus, much less impact long-term adaptive immunity. However, we are intentionally disrupting a beneficial immunological response with viral proteins, while infecting with a different, albeit harmless, virus (good old Sendai). Does this count?
Interpretation: depends a lot on how “beneficial immunological response” is defined
Possible effect: any research could be banned if it alters host responses categorized as “immunological”
(c) Conferring to the agent or toxin resistance to clinically or agriculturally useful prophylactic or therapeutic interventions against that agent or toxin or facilitating their ability to evade detection methodologies
This provision bans all molecular cloning. When we clone a gene, we generally put it in a plasmid that contains an antibiotic resistance marker. Then we put that plasmid in E. coli and plate it on media containing the antibiotic. The antibiotic allows us to select for our clones, since bacteria that don’t contain the clone with the resistance marker won’t be able to grow. Although lab strains of E. coli are not very dangerous, they are pathogenic. They can cause urinary tract infections and I’m sure can cause opportunistic infections if given the chance. E. coli does qualify as an “agent” and all cloning intentionally confers antibiotic resistance to it. Typical antibiotics are ampicillin, kanamycin, and tetracycline, all of which are clinically and agriculturally useful interventions.
Besides banning cloning, this also will ban any research into antiviral and antimicrobial resistance, including the development of new drugs. A critical part of studying antiviral and antibiotic drugs is to study how easily resistance might emerge. That involves screening to select for resistance mutants, which inherently means creating conditions for those mutants to emerge. Similarly, experiments to see if diagnostic tests can detect new variants of a virus could be prohibited as well.
Interpretation: could be inclusive of all molecular cloning, definitely is inclusive of most antimicrobial and antiviral resistance testing
Possible effect: severe restrictions of molecular biology capabilities, shutting down the antimicrobial and antiviral drug development pipeline
(d) Increasing the stability, transmissibility, or the ability to disseminate the agent or toxin
Finally, the GOFROC ban, now with “stability” added to the definition, and no exemption for any species. This is so broad that it effectively categorizes any experiment growing any RNA virus forbidden GOFROC. RNA viruses are very mutation prone because the enzymes that copy RNA make a lot of mistakes. This is an advantage for RNA viruses, because they can rapidly begin acquiring adaptive mutations, including in cell culture. Any time you grow up viruses in cells, they start to become more fit: replicating more efficiently and eventually producing more virus. Of course, this often makes the virus lose other features. SARS-CoV-2, for example, will lose its fabled furin cleavage site when it is cultured in Vero cells, which are the standard for growing them. This makes the virus less transmissible between hosts, even though it makes it better at growing in Vero cells. But would this be allowed, since it will result in higher titers in cell culture? Would you be able to grow RNA viruses—any RNA viruses at all? If some RNA viruses are exempt, where is the line? Would you be able to work with seasonal influenza viruses or common cold coronaviruses, or would they be considered too pathogenic? What about something like PR8, which is a lab-adapted flu virus that doesn’t really even infect people but was derived from a human case in the 1930s? It’s not a dangerous virus, but it is a pathogen and it follows the same evolutionary pressures as everything else. How can viruses be studied if growing them at all violates this policy?
This would also likely stop a whole bunch of vaccine development work in its tracks. One strategy for making vaccines, especially live attenuated vaccines, viral vectored vaccines, or old-fashioned inactivated vaccines like Taubenberger’s $500M Generation Gold Standard initiative, is to select for attenuating mutations that increase fitness and thus increase virus production. Even though attenuating mutations by definition reduce pathogenicity, this research could be banned on the grounds that it selects for more fit vaccine viruses.
Interpretation: could be inclusive of all virus culture or experimental infection studies in vivo
Possible effect: no more growing viruses. Any viruses. Attenuated vaccine development could be prohibited.
(e) Altering the host range or tropism of the agent or toxin
Host range and tropism refer to the species susceptible to infection and the cell and tissue types that get infected. Changing host range or tropism is very common in virology when it comes to developing models for studying pathogenesis (how the infection causes disease). Often viruses that cause human disease don’t infect mice very well or don’t cause disease in these animals. However, mouse models can be very useful for this type of work. Mice are small and easy to handle, so they are safer to work with in high containment, as well. One approach to using mouse models for some pathogens is to mouse-adapt them. This was a big part of my PhD research. I mouse-adapted a human rhinovirus, which causes the common cold, by passaging it through mouse respiratory epithelial cells repeatedly. This gave the virus the ability to infect mice (not very well!), so it changed the host range. However, it didn’t change the virus’ ability to infect human cells, and it certainly did not cause any disease in mice (to my great sadness, since I was trying to make a mouse model of the common cold). It’s a pretty weak human pathogen that I adapted to make it infect mice without making them sick, without modifying its properties in human cells. This work changes host range, but doesn’t make the pathogen more dangerous to any host. Should it be prohibited?
More broadly, the term “tropism” could do a lot of heavy lifting here for how general the effects of this rule could be. If they are referring to tissue tropism, that would likely only apply to experiments selecting for viruses that could be transmitted by different routes. But if this refers to cellular tropism, that would then apply to culturing any virus in a new or different cell type, which we do all the time. Rhinoviruses normally don’t grow in the cervical epithelium but I grew all my virus stocks in HeLa cells.
Interpretation: applies to any type of host adaptation. Potentially applies to all virus culture, if cellular tropism is included
Possible effect: no more growing viruses except in very rigidly defined experimental systems
(f) Enhancing the susceptibility of a human host population to the agent or toxin
As someone who spends a lot of time thinking about host susceptibility, I’m actually at a loss for what this provision means in any kind of good faith context. I have no idea how you would enhance the susceptibility of a human host population. Run around dosing everyone with corticosteroids or immunosuppressants? Try to CRISPR out interferon or something, at population scale? Covert gene therapy to make everyone a receptor transgenic? I seriously can’t think of a valid scientific justification for why someone would try to make a human population more susceptible to a given pathogen outside of a terrorist attack, and there are already multiple federal laws that unambiguously ban bioweapons production and bioterrorism.
While I can’t think of a legitimate scientific reason for this policy point, I can think of some unscientific reasons. A common anti-vax lie is the assertion that vaccines do not elicit immune protection and actually make people more vulnerable to infection. What happens when Secretary Kennedy decides that he doesn’t like a new vaccine under development? He doesn’t need to worry about vaccines that will never make it into the clinic because of false assertions that they might “enhance the susceptibility of a human host population.”
Interpretation: completely unrealistic unless used to support anti-vax claims that vaccines increase susceptibility
Possible effect: vaccine development pipeline dries up
(g) Generating or reconstituting an eradicated or extinct agent or toxin
Although I am aware of exactly zero ongoing virus de-extinction efforts anywhere, with or without NIH funding, I guess it doesn’t hurt to emphasize that the virus resurrection party ended with 1918. Granted, I’m not sure what viruses besides 1918 are even good candidates for our dark reverse genetic magicks—1918 was really the only obvious choice for virus necromancy since it’s the one that’s useful for studying the properties of a pandemic virus. But at least this is the one part of this policy that is clear about the scope of what it is banning.
Interpretation: no more virus de-extinction
Possible effect: no more virus de-extinction
My biggest objection to this policy is the potential for it to be applied imprecisely. Most of the policy provisions are prohibitions that leave little to no room for an accurate, evidence-based, quantitative risk assessment. There are no stipulations that expertise is required to understand context, which is essential if this policy is to be applied with the intention of preserving necessary research and effectively mitigating risks.
To my knowledge, Nickels and his fellow lab leakers on the policy implementation team, Alex Washburne and Ed Hammond, have no technical experience with virology research. In fact, they have no expertise in virology at all. Nickels is a former professor of genetics who studied bacterial transcription before becoming an anti-virology zealot, Washburne is a self-proclaimed Renaissance man and insufferable pick-me who claims expertise in everything from short selling pandemic stocks to designing exceptionally silly fantasy coronavirus cloning scenarios, and Hammond is a professional curmudgeon and cherry-picking public records gadfly. What they and Bhattacharya have all amply demonstrated, along with their credulous ignorance about molecular virology, phylogenetics, and pathogenesis, is their hostility and bias toward an entire professional field. That does not bode well for these policy experts taking a thoughtful or objective approach to implementation.
As a result, I have no confidence that this policy will be implemented in a way that fairly assess research for its risks and benefits. I expect these policies to be implemented selectively and punitively, and specifically weaponized against scientists and research that the administration wants to defund. This includes basic research on diabetes, cancer, Alzheimer’s, fertility, aging, nutrition, autoimmune disease, and genetic disorders. This includes developing new vaccines and drugs. This includes meaningfully assessing the risks we face and developing our ability to reduce them. This is not just virology research. This is everything. Most of the work that will be banned will not actually be dangerous in terms of pandemic or public health risk. It will be dangerous to the ideologues who are attacking science as a means of enforcing authoritarianism upon us, which is why it will all become “GOFROC.”
The American people need to understand that this is all based on a lie about where the true danger. There are dangers in virology labs, although these can be mitigated and risks greatly reduced. There are also dangers in nature: millions of potential pathogens, waiting to emerge. There are so many dangers in nature that we don’t know much about most of them. The only way to mitigate these dangers is by studying them in a safe and responsible way, so we know how to reduce the risk they present. This policy will prohibit doing exactly that. It is a biosafety policy that will measurably increase danger and reduce safety. It’s a gain of fiction, and it will be all of our loss.
this is a detailed and thoughtful analysis of one small aspect of the travesty being wrought by donOLD's administration in public health and research
Keep calling out the lies by the Trump Administration. This is a very informative article about your work and its importance to world health and I will spread your information in my circle. Thank you. Lead on.