Coronavirus biosafety levels

"On reading," by Simon Wain-Hobson, is a weekly discussion of scientific papers and news articles around gain of function research in virology.

Since January 2024, Dr. Wain-Hobson has written weekly essays for Biosafety Now discussing risky research in virology. You can read his entire series here.

Spring azaleas

On reading Bat-infecting merbecovirus HKU5-CoV lineage 2 can use human ACE2 as a cell entry receptor by Jing Chen and colleagues, Cell S0092-8674(25)00144-8; Recent research should raise alarm, by Ralph Baric and Ian Lipkin, a guest essay New York Times, March 3, 2025; We were badly misled about the event that changed our lives by Opinion columnist Zeyned Tufekci, New York Times, March 16, 2025.

---

This scientific paper from the Wuhan Institute of Virology (WIV) represents a huge amount of work. As it has been the subject of discussion in the NY Times we’ll take a look at it first, then the comments.

The bat samples were collected in 2004. They identified five samples positive for a coronavirus by genetic screening methods and determined the sequences of their viral genomes. The were close cousins and distinct from the closest known relative. All were part of the MERS family of coronaviruses that first came to the fore in 2012 in Saudia Arabia. They showed that the viral Spike protein could use the human ACE2 receptor which allows docking of the virus to the cell and subsequent infection. This is precisely the same receptor used by the SARS2 virus. Interestingly the Spike protein could dock most mammalian and bird ACE2 lookalikes as well as multiple bat equivalents. All this was done by exploiting genes not the virus.

That a virus can dock with proteins from many species is not new. Think of SARS-CoV-2 that spread to deer, dogs and all sorts of cats. Or cowpox virus that protected milkmaids from smallpox. Such viruses are said to be amphotropic, that is they can infect more than one species.

They then isolated a virus, called HKU5-CoV-2, on human cells in the lab. Thereafter follows a festival of molecular structural biology all of which comes to the conclusion that while belonging to the MERS group of coronaviruses, which usually dock with the DPP4 molecule, the Spike protein of HKU5-CoV2 recognizes the human ACE2 molecule in a manner similar to the human SARS-CoV-2 virus. All this is nice to have, good to know.

Work with the HKU5-CoV-2 was performed in a BSL-2 plus negative pressure following SOPs with necessary personal protection and approved by the Wuhan Institute of Virology (WIV) IBCs (Institutional Biosafety Committee; SOP standard operating procedure, BSL-2 Biosafety Level 2).

If you check out the 2024 NIH guidelines (Appendix B-V) you can find the following: For agents that are infectious to human cells, e.g., amphotropic and xenotropic strains of murine leukemia virus, a containment level appropriate for RG2 human agents is recommended. This means BSL-2 labs. (A xenotropic virus is one that cannot replicate inside the animal it was found in but can infect a different species. Yes! Biology is varied and wonderful).

The same document says that coronaviruses can be handled in a BSL2 lab. The only exceptions are the SARS and MERS coronaviruses where BSL-3 labs are required.

Drs. Ralph Baric and Ian Lipkin, two excellent US virologists, commented on this work in an opinion piece in the NY Times. They had no criticism of the work whatsoever; their beef was with the biosafety level at which the work was done. There’s a central question that many scientists face: How can scientific discoveries drive humanity’s progress without posing a dire risk to it? …But we are concerned about how some scientists are experimenting with viruses in ways that could put all of us in harm’s way. High road rhetoric.

The Chinese group was following their own institutional recommendations which were comparable to USG recommendations.

In the WIV paper we read that During the preparation of this manuscript, two independent teams led by Professors Lekto and Veesler, respectively, released their results that lineage 1 HKU5-CoV (here referred to as HKU5-CoV-1) also use ACE2 orthologs from only a few species including their host, Pipistrellus abramus (P.abr) bat as receptor.30,31 However, neither study reported that these HKU5-CoV-1 can efficiently utilize human ACE2 for cellular entry.

The numbers 30 and 31 refer to two manuscripts on the same subject posted on prepublication websites as is common these days. That from Prof. Letko was posted on August 16, 2024 and has Ralph Baric as a senior author. The second was posted on August 28, 2024 from Prof Vessler’s group at the University of Washington in collaboration with labs in Riyadh and Wuhan (not the WIV).

The Letko/Baric work was presumably performed in the enhanced BSL3 lab at their disposal and has not yet been published in a peer-reviewed journal.

The UWash experiments related to wildtype authentic HKU5-1 infection were authorized by the Biosafety Committee of the Center for Animal Experiment of Wuhan University and conducted in a certified Biosafety Level 2 laboratory according to standard operating procedures (SOPs) in a BSL-2 laboratory. The virus work was performed in China. This work was published in the journal Cell on February 5, 2025, two weeks before the paper from the WIV discussed here.

The odd thing is that this paper is ignored in the Barib/Lipkin opinion in the NY Times, even though the UWash performed their virus work in a BSL-2 lab, the very biosafety lab they criticized. This is a deliberate omission for they knew of these preprints since August 2024. Why go for the WIV and not the UWash group?

There is no point speculating, but it weakens their NYT opinion. A pile on by Dr. Tufecki in the same newspaper followed 2 weeks later. Five years after the onset of the Covid pandemic, it’s tempting to think of all that as ancient history. We learned our lesson about lab safety — and about the need to be straight with the public — and now we can move on to new crises, like measles and the evolving bird flu, right?

Wrong. If anyone needs convincing that the next pandemic is only an accident away, check out a recent paper in Cell, a prestigious scientific journal. Researchers, many of whom work or have worked at the Wuhan Institute of Virology (yes, the same institution), describe taking samples of viruses found in bats (yes, the same animal) and experimenting to see if they could infect human cells and pose a pandemic risk.

This was no doubt influenced by Drs. Baric and Lipkin opinion as she too missed the earlier paper in Cell. Furthermore, Dr. Tufecki gives the impression the Chinese group was doing GOF 2.0 research which they were not. To inform public health people and planners, virologists must know just how abundant amphotropic bat coronaviruses are in the wild, particularly those with the ability to infect humans. That is the whole point of the work! If it was just another study of a bat specific coronavirus Cell would not have the paper for publication.

They may have been following local guidelines, however, what if the virus had adapted very quickly to human cells, much like Canine Distemper virus (More GOOFing), then what? Given the GOF 2.0 controversy it would have been wise to have done the work using human cells in a fully-fledged BSL-3 lab and perhaps an enhanced BSL-3 lab as it is probably a respiratory virus. On top of which, the WIV virologists knew before isolating the virus on human cells that the Spike protein of HKU5-CoV-5 could engage human ACE2 expressing cells.

Dr. Tufecki missed what the WIV was doing and the biosafety issue. The NY Times foghorn misused once again (Chilly New York times).

Drs. Baric and Lipkin didn’t miss the biosafety issue. They were right to point this right. The USG policy on DURC was, until mid-2024, anchored in attaching a biosafety level to work based on the virus going into the experiment. Certainly, growing an amphotropic bat coronavirus on bat or mouse cells is BSL-2 level work. However, growing it on human cells? Potentially a respiratory transmitted virus for humans?

Furthermore, it is very likely that growth on of HKU5-CoV-1 on human cells would have resulted in a few adaptive mutations. Such mutations have been observed so many times for a vast array of different microbes when grown in the lab, so it is safe to consider it as a given.

Drs. Baric & Lipkin move on. Scientists and policymakers in the United States have spent years discussing and debating how to regulate risky virus research, sometimes contentiously.

Agencies inside and outside government that fund this sort of work should require proof that investigators meet global standards. Scientific journals should have similar standards for the studies they accept.

Last week was the 50th anniversary of the 1975 Asilomar Summit, where scientists came together to establish guidelines for research with genetically modified microbes. Today many more discoveries and threats are on the horizon. Potentially dangerous research should not be done without proper precautions to prevent deliberate or accidental spread.

• The GOF 2.0 controversy revealed strong pushback to comments from scientists who dissented from the NIH position (Chilled virology). The American Society for Microbiology has been especially active pushing back against any form of external oversight (Crossing the line).

• Funders must be sure that the research they enable is performed in accordance with the biosafety levels of the place and day. Global biosecurity shouldn’t be left to a local review board to decide on a colleague’s work.

• Scientific journals shouldn’t have a word to say for 1) it is too late in the day, 2) the manuscript was probably uploaded to a preprint server, and 3) they have shown themselves to align with the establishment. They are not independent.

• The celebratory Asilomar conference can be forgotten as it dodged the issue of GOF 2.0 virology (Spineless again).

Conclusion

The biosafety level required for research on amphotropic viruses needs revisiting fast. That for novel animal respiratory viruses that can infect human cells, needs to ratchet up a notch.

---

Aside 1

Amphotropic, from Greek where ‘ampho’ denotes both kinds and ‘tropic’ affecting or attracted to the thing specified, in this case cells of an organism.

---

Aside 2

There is a huge difference between studying natural amphotropic bat coronaviruses and deliberately making a novel human virus along the lines of the Fouchier and Kawaoka H5N1 viruses. The former are in the here and now and need virologists’ attention, the latter are science fiction.

---

Aside 3

The WIV and UWash studies are good to have. True, public health planners can’t act on the information easily as pandemic preparedness is a very tough issue (Following the science). However, virologists could

• find out just how easy, or difficult, it was to make vaccines to these beasts. Could vaccination protect against related viruses?

• find out if existing antivirals were of any consequence. The WIV group showed they were. These two bullet points are part of continuing surveillance.

Next time round, what if someone found an amphotropic bat coronavirus that grew super well on ACE2 positive human cells which was not the case here?

• Hopefully the experiment would be performed in a BSL-3 lab.

• That someone would probably utter a four-letter word a few times and pick up the phone. OK the virus might not yet have gone viral, if you’ll forgive the pun, but hell.

More good reasons to support infectious disease research.

---

Aside 4

As the HKU5-CoV-2 virus was isolated from a sample collected in 2004, the WIV has a large number of samples that they are slowly working up. Good for them. It easy to say with hindsight, but if this had been known before the 2012 MERS outbreak in humans it might not have come as such a surprise.

Yet we’ve seen this before with the bat coronavirus RaTG13. Its sequence was published in March 2020 just after the start of the COVID-19 pandemic. It was derived from a bat fecal sample retrieved from a cave in Yunnan Province linked to an outbreak of pneumonia in 2012 among miners who collected bat guano. Samples were collected from the cave between August 2012 and July 2013. PCR data for this virus, referred to as 4991, was known by 2016. In short, a relative to the COVID-19 virus was known to us before the pandemic. Again, if we had known this sooner, SARS-CoV-2 would have been a little less of a surprise.

The depth (to use a technical word) of the family (genetic) trees relating SARS-CoV-2 like bat viruses suggests vast numbers of distinct coronaviruses. It would be good to have many more of these undocumented relatives isolated. Yet as mentioned in the essay We the virologists, there is resistance to doing this.

---

Aside 5

Without a microbiology background you might think Dr. Tufecki could be granted attenuating circumstances, just as science journalism can be so hard (Some housekeeping). She could have simply refrained from writing about stuff she doesn’t know much about or ask someone who does. As a regular opinion writer to the widely read NYT Dr. Tufecki must be held to high standards.

---

Aside 6

It will not have escaped the reader’s attention that Dr. Ian Lipkin is an author on the notorious Proximal origins of COVID virus paper. Notorious as it made claims that had no scientific data to back them up. Notorious when the inside story was exposed by numerous investigators around the world. Despite these failings it had huge impact effectively shutting down discussion of the lab leak hypothesis for a year or more.

This doesn’t mean that all from Dr. Lipkin, who has distanced himself from this paper, is comparable. Far from it, he’s a creative and solid virologist.

The upshot is if you screw up - and nobody’s perfect - say so and move on. It’s a sign of strength. Colleagues get that.

---

Simon Wain-Hobson is an emeritus professor at the Institute Pasteur, Paris, from which he retired in 2021. He and his colleagues were the first to sequence the genome of HIV, and Wain-Hobson has published more than 230 papers on virology and cancer.