13.
Accident
‘But I’ve never seen the Icarus story as a lesson about the limitations of humans. I see it as a lesson about the limitations of wax as an adhesive.’
RANDALL MUNROE
It is now the turn of the attorney for the other side to make the case that this pandemic derives from an accidental laboratory leak. We begin by making clear that we are not alleging malfeasance, only a mistake. The hypothesis that the Chinese government deliberately engineered a virus and released it to cause harm in the world is not the case that is being argued.
Rather, the allegation we take seriously is that scientists at the Wuhan Institute of Virology and, possibly, other laboratories in the city, at the Center for Disease Control or the Huazhong Agricultural University, were doing exactly what they said they were doing: namely, studying viruses from bats and other wildlife with a view to predicting pandemics and eventually developing therapeutics or vaccines. But something went wrong and a virus leaked, either infecting a person who worked in the field, in a laboratory, or elsewhere in the city if lab waste had not been properly decontaminated before being discarded.
The first point is that such research accidents, while rare, are sufficiently frequent that they can be expected to happen. Lottery wins are extremely rare, but they happen every week. If a lab works with tens of thousands of diverse zoonotic virus samples over a period of several years – especially at lower biosafety levels – it is only a question of when, and not if, a biosafety incident occurs. Unfortunately, there are no enforceable international biosafety and biosecurity standards. Many accidental releases of SARS, anthrax, smallpox, foot-and-mouth, Marburg virus, and other pathogens have occurred and continue to occur in even the most secure and well-run laboratories. Remember the SARS laboratory leaks in 2003 and 2004. One researcher had embarked on an international trip, and another researcher had taken multiple long-distance train rides after getting infected with SARS virus in the laboratory. For the former, it was pure luck that he had not developed symptoms and become infectious while in international transit. For the latter, close to a thousand people were quarantined and her laboratory-acquired infection had only been discovered a month after the fact.
Next consider that there is no direct evidence of a natural origin of the Covid-19 pandemic. As each month goes by, it becomes more and more extraordinary that the proponents of a natural origin fail to find any of the sort of evidence that very soon came to light in the case of the 2003 SARS epidemic. In that episode it quickly became apparent that food handlers were over-represented in the early cases, which led to the testing and discovery of infected animals in markets. Animal traders in Guangdong were found to have antibodies against SARS viruses despite not having been previously diagnosed with SARS or displaying symptoms – meaning that these groups of people had been regularly exposed to SARS-like viruses prevalent in the animal trading circuit. Guangdong was a hotspot for SARS viruses to transmit into humans. True, it then took months to find bats carrying similar viruses and then years to find the original bat reservoir of very closely related viruses. But the market evidence was strong and came early. Although the technology at that time was limited compared with today’s, the intermediate hosts of the SARS virus were identified within weeks of first sequencing the genome of the virus. And when SARS spilled over once again at the end of 2003, Chinese doctors and scientists were even swifter at finding infected civets at the restaurant workplace of an index patient – on the very day of the patient’s SARS diagnosis, the civets were sampled. No time was wasted.
This time, despite testing markets, farms, and eighty thousand animal samples spanning dozens of species across China, no similar evidence has emerged for SARS-CoV-2. Hundreds of samples taken from the animal carcasses at the Huanan seafood market all tested negative for any trace of the virus. A considerable fraction of the early Covid-19 cases in Wuhan also had no exposure to the market. Damningly, one of the early lineages of SARS-CoV-2 was not detected at or associated with cases from the market – pointing to the more likely scenario that a person had brought one of the early variants of the virus into the market where a superspreading event occurred in a poorly ventilated and crowded venue.
In stark contrast to the 2003 SARS virus, there has been no explanation of how it came to be so adept at spreading among humans. Both the 2003 SARS and SARS-CoV-2 are generalist viruses and can infect a wide range of animal species, but the stepwise adaptation of the 2003 SARS virus to its new human host in the early months of that outbreak is missing altogether in the case of SARS-CoV-2. When it was first detected in Wuhan, it had apparently stabilised genetically – there was just a single mutation, D614G, in the early cases that could be said to slightly enhance its transmissibility among humans. More variants came along later, of course, as the virus became immensely widespread, increasing the opportunities for mutation in millions of victims. If the 2003 SARS virus had been allowed to spread to millions of people, we would have certainly seen the evolution of variants – even after the virus had picked up dozens of helpful adaptive mutations in the first months of its exploration in human beings. In the case of SARS-CoV-2, the Alpha variant has been hypothesised to be the product of long-term infection of an immunocompromised individual. The Delta variant can re-infect survivors of Covid-19 and emerged in a part of the world where Covid-19 was running rampant. In the absence of any intermediate host or remote human populations found to carry SARS-CoV-2, where did the virus perfect its ability to infect and transmit among humans? Could it have acquired its skill in a lab, dwelling in the cell cultures derived from the human respiratory tract, or even in humanised mice?
Ask yourself this: why did the Chinese authorities refuse to release to the WHO investigators the raw data about the early Covid-19 cases in Wuhan? If they wanted to quash rumours of a laboratory leak, the simplest thing to do would have been to share unredacted hospital records from Wuhan with outside investigators, alongside information on the locations and professions of the first cases. If these showed no connection to any virus laboratory in the city, it would argue against a laboratory leak. Yet no such details have been disclosed.
Ladies and gentlemen of the jury, says the attorney, let me pose a very simple question: why Wuhan? Of all the gin joints in all the towns in all the world, she walks into mine, says Humphrey Bogart’s Rick in the film Casablanca when Ingrid Bergman’s Ilsa turns up in his bar. But if you were escaping Nazi-occupied Europe, Casablanca was a well-known rat run with regular flights to Lisbon, and Rick’s Bar was supposedly the main place to go in the town to buy travel papers, so it’s not quite as improbable as he makes it sound. It is surely far stranger that a bat sarbecovirus should walk into the very city with the largest collection of bat sarbecoviruses in the world and the most active research programme studying such viruses – and a city that is well over a thousand kilometres from the region where closely related viruses are found naturally.
Dr Peter Daszak and Dr Thea Fischer, of the WHO team tasked with investigating the origins of Covid-19, arriving at the Wuhan Institute of Virology, 3 February 2021.
REUTERS/Alamy Stock Photo
Scientists from Wuhan regularly made the long trip to southern China and trekked through forested terrain to reach remote caves in order to collect their thousands upon thousands of samples for study back in the metropolitan city of Wuhan. They did not always wear full protective gear as they handled thousands of bats, swabbing the animals’ anuses and noses and generally getting far closer to them than local people ever would. It is highly unlikely that any other route to Wuhan was so conveniently available to such viruses. Even among the villagers living near caves in which SARS-like viruses had been found in bats, only 2.7 per cent had antibodies against SARS. Of the six individuals who were seropositive in that study, only one had left the province in the past year, and two had not even left the village. This is how remote these villages are. A bat virus with a lust for travel might wait decades to get a chance to visit nearby Kunming, let alone distant Wuhan. From the scientists’ own extensive sampling of bat viruses across multiple Chinese provinces over the years, a simple truth emerges: within China, the particular lineage of bat sarbecoviruses that SARS-CoV-2 belongs to was only ever found in Yunnan province – certainly nowhere close to Wuhan.
We now know, as we did not in February 2020, that the WIV was in possession of a batch of very closely related viruses collected from an abandoned copper mine in Mojiang where workers had sickened with a SARS-like illness in 2012. We also now know that the WIV had not been forthcoming with vital information about these viruses, including what had been done with them in the laboratory. We know that for months the fact that more than one such virus had been collected from the mine went undisclosed. If not for the chance event that a large batch of sequences was uploaded with a scientific paper to an international genomic database before the pandemic began, the existence of eight of these viruses might never have been uncovered. And if not for the efforts of the sleuths and scientists painstakingly sifting through the raw data and supplementary materials of scientific papers, comparing sequences and compiling spreadsheets, the link between the Mojiang miners and the SARS-CoV-2-like viruses under study at the WIV might have never been brought to light.
The search for a virus that is a closer match elsewhere than Mojiang has been a failure. Some proponents of the natural-origin theory greeted with enthusiasm the discovery of viruses related to SARS-CoV-2 in Thailand, Cambodia and Japan, but none was as close a match to SARS-CoV-2 as the WIV’s RaTG13 from the mine. Every new discovery of a less closely related virus serves only to underline the significance of the viruses brought from the Mojiang mine.
Then there is the almost ridiculous secrecy that surrounds the mine itself. To this day Dr Shi and her colleagues have refused to confirm the exact location, though it is no longer a secret thanks to the various theses and the work of sleuths. The spot remains strictly off limits to foreigners and is heavily guarded. International journalists trying to visit the Mojiang mine have been tailed, obstructed by supposedly broken-down vehicles, and detained by the police. This is not an approach that instils confidence in the innocence and irrelevance of the site.
Pull back from the detail and focus on the simple fact that the WIV had a database of at least fifteen thousand bat samples taken mainly from southern China. With more than twenty-two thousand entries in total, that database included the dates and locations of samples and the descriptions of viruses found in them. It became inaccessible to users outside of the WIV in September 2019 and was taken down altogether sometime in early 2020. To dispel rumours that SARS-CoV-2 was derived from this collection of virus sequences or samples, the WIV could have easily shared this database with other scientists. The excuse that there had been ‘hacking attempts’ does not make much sense, because that would not prevent the sharing of the data with other scientists. Yet, well over a year later, no one has reported having a copy or access to this database. What’s the point of collecting viruses if you hide the data when a pandemic actually occurs?
The next piece of evidence against the laboratory scientists is the long and detailed record of their research. In paper after paper, they laid bare a record of experiments on sarbecoviruses and other coronaviruses that were ingenious, comprehensive and successful. They did not just bring viruses from caves and mines in southern China to the laboratory for storage; they sequenced their genomes, made infectious clones of them, rescued live viruses from culture, passaged them through a range of laboratory-made cell lines of different animal species and cell types, synthesised and altered their genomes to insert specific sequences, hybridised genomes to combine parts of one virus’s spike gene with backbones from another virus, and used these viruses to infect human respiratory tract cells and humanised mice genetically engineered to have human ACE2 in them. This type of research carries a risk of, unintentionally, generating a more virulent or infectious version of a virus or selecting for bat viruses that are efficient at infecting human tissues and humanised animals. Their purpose was admirable: to understand the risk that each newly discovered virus posed and perhaps one day to devise a vaccine against all SARS-like viruses. But they also brought a possibility of starting a pandemic.
Nor can we consider fatal the argument made by proponents of a natural origin that the virus has not been reported in the scientific literature and shows no traces of genetic engineering. Scientists know that it often takes years to build up a solid scientific publication. Discoveries are not reported in real time. Samples are stored for years, awaiting analysis. We also know from the Washington Post that there were secret projects at the WIV and some reports or theses could be sealed for up to two decades. The WIV database is a case in point: to our knowledge, the existence of many of the entries has not yet been reported in formal scientific papers. The 630 viral samples reported in the Latinne et al. paper in 2020 had all been collected from 2010 to 2015 yet nobody thinks the WIV stopped collecting and isolating viruses in 2015. Besides, RaTG13 itself was collected in 2013, sequenced fully in 2018, but not published until 2020 when it became imperative to clarify its similarity to SARS-CoV-2. The other eight viruses collected in 2015 from the Mojiang mine were undisclosed until 2019 and their sequence was only released in mid-2021.
In any case, a laboratory leak does not require the pathogen in question to have been genetically modified. Many laboratory escapes have involved viruses collected from naturally infected animals or human beings. But if an engineered virus does escape, there might be no way to distinguish it from a natural mutant. Today’s technology allows the seamless construction of entire virus genomes. We know that the WIV was collecting thousands of natural virus specimens, sampling from thousands of animals and people living in rural areas. We know that they were using and improving on these seamless techniques for switching parts in and out of virus genomes. And we know that they were testing the novel SARS-like viruses, natural or chimeric, in human cells and in humanised mice and civets. These conditions would have selected for a largely natural virus that could effectively use the human ACE2 receptor and perhaps spent time inside animals with functioning immune systems. These laboratory processes would not have left an obvious mark of laboratory origin on the virus.
The significance of the S1/S2 furin cleavage site in the spike gene of SARS-CoV-2 is unclear. On the one hand, such sites appear in many viruses, including other coronaviruses. On the other hand, it is striking that no other sarbecovirus has one, including all of SARS-CoV-2’s closest relatives and the dozens of other SARS-like viruses collected from bats. To us, however, the most shocking aspect of the furin cleavage site is Dr Shi’s failure to mention it in early papers first describing the novel coronavirus. The furin cleavage site stands out like a sore thumb and is a large part of the reason this virus is so infectious and virulent. In one recent collaboration between Dr Shi’s group and scientists in America, a furin cleavage site had been engineered into the spike of a bat MERS-like coronavirus. So there is no doubt that Dr Shi and her colleagues understood the functional significance of this feature in coronaviruses. Yet the Nature paper co-authored by seasoned coronavirus specialists diligently described other minor features of the spike gene sequence but missed the show-stopping furin cleavage site. It was left to other teams of scientists in China, Canada and France to say ‘Hey, look – this is the first sarbecovirus with a furin cleavage site, which might explain why it is so highly infectious.’ Across town from the WIV, scientists at another Wuhan research institute, the Huazhong Agricultural University, had also recently collaborated with Dutch scientists to insert such a site into the spike of a live porcine epidemic diarrhoea coronavirus. It is a remarkable coincidence at the very least that the first sarbecovirus with a furin cleavage site showed up in one of the few gin joints in all the world where the knowledge and capabilities exist for inserting one into a coronavirus.
One final argument for the laboratory leak is that more virologists have shifted their view and now either think it is possible or even likely. Given the potential risk to the reputation of science as a whole and virology in particular, this cannot have been easy for them. Dr Ian Lipkin of Columbia University was a co-author on the ‘Proximal Origin’ paper that had stated in March 2020: ‘We do not believe that any type of laboratory-based scenario is plausible.’ However, in May 2021, Dr Lipkin said of the experiments done at the lower BSL-2 biosafety level at the WIV: ‘That’s screwed up. It shouldn’t have happened. People should not be looking at bat viruses in BSL-2 labs. My view has changed.’ Dr Bernard Roizman, a virologist at the University of Chicago, who signed the Lancet letter drafted by Dr Peter Daszak in February 2020, went further. In May 2021, he said: ‘I’m convinced that what happened is that the virus was brought to a lab, they started to work with it . . . and some sloppy individual brought it out . . . they can’t admit they did something so stupid.’ Another author of the Lancet letter, Dr Charles Calisher, a Colorado State University virologist, also told ABC News that ‘it is more likely that it came out of that lab’. And a third author of the letter, Dr Peter Palese, told the New York Post that ‘A lot of disturbing information has surfaced since the Lancet letter I signed, so I want to see answers covering all questions.’ Dr Gary Whittaker, a specialist in influenza and coronaviruses, joined Dr David Baltimore in calling the furin cleavage site in SARS-CoV-2 ‘extremely unusual, leading to the smoking gun hypothesis of manipulation’.
It must be conceded that many of those making the charge that SARS-CoV-2 leaked from a laboratory are not agreed as to which of several versions of the leak they find most plausible: whether SARS-CoV-2 is a natural virus that was stored there and leaked, like the SARS leaks in Beijing in 2004; or a virus that had been passaged in human cell cultures to the point where it evolved into an efficient human pathogen; or whether SARS-CoV-2 is an engineered chimeric virus. All are possible and until there is better evidence none of these can be ruled out.
Ladies and gentlemen of the jury, the attorney concludes, there is a stark absence of evidence for a zoonotic spillover at the start of this pandemic. In contrast, the proximity of the outbreak to the WIV – the largest collector of SARS-related coronaviruses in the world, where scientists were creating chimeric viruses and experimenting with close relatives of SARS-CoV-2 – makes a compelling case for a laboratory-based origin of the virus.