1.

The copper mine

‘The very cave you are afraid to enter turns out to be the source of what you are looking for.’

JOSEPH CAMPBELL

In the spring of 2012, six men were admitted to a hospital in Kunming, the capital of Yunnan province in south-west China. The chief symptoms were dry coughs, shortness of breath, high fevers, aching muscles, headaches and fatigue. All six had recently worked in the same mine in Mojiang County, clearing out bat guano, up to 150 metres deep in the bat-infested, man-made cave. The four oldest patients became critically ill and suffered respiratory failure; three eventually succumbed to the mysterious disease and died. There were signs that their immune systems had been severely damaged, allowing for opportunistic infections. This, in combination with other clinical diagnoses, suggested that an unknown viral infection was highly likely to be the cause of their affliction.

The first patient, 吕 (Lu; the full names of the patients were obscured in the thesis), aged forty-two, was admitted to the hospital on 25 April and died on 12 June. The oldest patient was admitted on 26 April and had the surname 周 (Zhou). Zhou was sixty-three years of age and died on 7 May. The two other patients also admitted on 26 April, 刘 (Liu, aged forty-six) and 李 (Li, aged thirty-two), both survived the ordeal, albeit Liu struggled in the hospital for months and was only discharged on 10 September. A fifth patient, 郭 (Guo, aged forty-five), was admitted on 27 April and died on 13 August. The last patient, 吴 (Wu, aged thirty), was admitted on 2 May. Both Li and Wu, only in their early thirties, were discharged on 28 May. The less time the patient had spent in the mine, and the younger they were, the better their prognosis and the shorter their hospital stay.

The outbreak caused alarm, and the attending physician noted afterwards that, if future cases of severe pneumonia were to be encountered in the hospital or clinic, it would be necessary to be alert to the possibility of infectious disease and take precautions against transmission in the hospital. In what sounds like an increasingly desperate attempt to diagnose and treat the cause of the sickness, which failed to respond to a barrage of antibiotics and antifungals, the doctors tested the patients for HIV, cytomegalovirus, Epstein-Barr Virus, Japanese encephalitis, haemorrhagic fever, dengue, Hepatitis B, SARS and influenza.

By the start of June, the oldest patient had died and the two youngest patients had been discharged. Senior medical experts were consulted for the remaining three critically ill cases. On 4 June, the hospital consulted Dr Xie Canmao of Sun Yat-sen University’s Department of Respiratory Medicine. He thought there was a ‘great possibility of fungus infection’; however, two more patients, Lu and Guo, died later in June and August respectively despite the administration of antifungal therapy. On 19 June, Dr Zhong Nanshan, also of Sun Yat-sen University, was consulted on Guo and Liu, by then the two remaining patients in the hospital, and came to a quite different conclusion: ‘great possibility of virus infection’. Dr Zhong is well known in China as one of the heroes of the Severe Acute Respiratory Syndrome (SARS) epidemic of 2002–3. Born in 1936 in Nanjing, Dr Zhong trained in Beijing and at Edinburgh University Medical School, where he obtained his medical degree in 1981. He was working at the Guangzhou Institute of Respiratory Diseases in 2002 when the SARS epidemic began. It was Dr Zhong who insisted, at some risk to his own reputation, that the disease threatened a major pandemic, and he subsequently devised a treatment, based on cortisone and oxygen, that saved many lives.

For both patients Dr Zhong recommended: one, identify the type of bats at the mine; two, test the patients for SARS virus and antibodies; three, treat them with a series of antifungals and antibiotics; and four, increase airway management and apply bronchoscopy for sputum suction. Sadly, Guo could not be saved and died on 13 August. Liu survived. In May, he had been treated with antithrombotic therapy (preventing blood clots) and showed significant improvement two days later. The doctors continued the anticoagulant treatment until he was discharged in September, more than four months after he had been admitted to the hospital.

The saga of the sick miners, potentially infected by a SARS-related coronavirus in a bat-infested mine, did not go unnoticed by prominent laboratories in China. This outbreak was of such import that as well as Dr Zhong Nanshan, it drew in the Wuhan Institute of Virology (WIV), the Chengdu Military Center for Disease Control, the Beijing Institute of Pathogen Biology and even the laboratory of Dr ‘George’ Fu Gao, the deputy director of China’s national Center for Disease Control and Prevention. (The CDC is a network of regional public health laboratories throughout China with headquarters in Beijing. Dr Gao was promoted to director in 2017.)

An abandoned copper mine

Six hours’ drive south of Kunming, not far from the border with Laos, lies Mojiang County. It is designated by the Beijing government as an autonomous county, in recognition that its indigenous inhabitants, the Hani people, are a distinct ethnic group. The area is hilly, heavily wooded and sparsely populated, but with terraced fields on some of the hillsides, accessed by switchback dirt roads. The terraces have been used for growing bananas, rubber, tobacco and tea, but dense, green vegetation cloaks many of the slopes. The biggest city in the area, Pu’er, has long been famous for its tea plantations and the dark, fermented tea made from them. East of Pu’er, about twenty kilometres south of the small town of Tongguan, in the partly wooded terrain on the left bank of the Babian river, a small creek called Bengpinghe leads up into the hills. On a ridge to the south of this creek, surrounded by groves of orange trees, stands a tiny hamlet called Danaoshan. A short distance from here are the remains of an abandoned copper mine.

The location of the mine where the six miners had worked has never been officially confirmed. However, relentless digging into Chinese databases by a group of diligent sleuths unearthed a 2016 doctoral thesis, this time from the laboratory of the deputy director of the Center for Disease Control and Prevention in Beijing, Dr George Gao, which identified the mine’s precise location: N 23°10’36’ E 101°21’28’. As revelations about the Mojiang miners and their potential connection to SARS-CoV-2 spread on Twitter in the second half of 2020 and early 2021, a growing number of journalists took it upon themselves to visit the mine, and each faced impromptu roadblocks, official excuses and local people deterring them from getting close to it.

Tongguan Township in Mojiang County, Yunnan, September 2018.

Zhou Lei/Xinhua/Alamy Live News Xinhua/Alamy Stock Photo

In October 2020, using the GPS coordinates from the thesis, the BBC’s John Sudworth and colleagues attempted to drive to the mine from Kunming. According to Mr Sudworth, they were ‘followed constantly for hours by as many as half a dozen unmarked cars’. As they approached from the east, the road became impassable, so they got out of their car and tried to reach the site on foot. After a long hike on rough ground, they got to the village of Danaoshan, but they were being watched. A man they met on the path refused to speak. By the time they arrived at the area the coordinates pinpointed, it was pitch black and they had to return and try again the next day. Unfortunately, their first attempt had raised the alarm and ‘the authorities were more than ready’ for their second visit. Mr Sudworth and his team tried again by car from a different direction but encountered a red construction truck blocking the road. They outsmarted the plain-clothes police following them by squeezing past the truck ‘Houdini-like’, with millimetres to spare, only to find another roadblock ahead. They set out on foot again but were ‘intercepted by some very angry men, in a 4×4, clearly communicating with someone in higher authority’. Groups of men were hanging around the area, and Sudworth was warned that ‘they would soon turn violent’ if he did not leave. The BBC team backtracked and tried a different route but were blocked again and again. One man told them that his job was to keep them out and that they would not be able to enter Danaoshan village again. Sudworth gave up on the Mojiang mine but tried to visit a nearby bat cave, named Shitou cave, in Jinning County, where scientists had found viruses most closely matching the 2003 SARS virus. They were met by another lorry blocking the road and men in military uniforms. They were held in a field for more than an hour before being forced to leave. ‘By now you get the idea. It’s impossible to overstate just how large and coordinated the effort was – state-security, plain-clothes police, uniformed police, officials and local residents. When we tried to talk to anyone, they’d turn their backs,’ Mr Sudworth reported in May 2021, by which time he had relocated to Taiwan.

It was clear that efforts were being coordinated to stymie journalists trying to retrace the steps of the virologists who visited the mine, and John Sudworth was only one of many thwarted reporters. A team from NBC’s Today show were told that wild elephants were on the road so they could go no closer. In 2021, a team of undercover French journalists got close to the site and managed to speak to somebody in Danaoshan. Asked if there was a mine nearby, he replied, ‘Yes, the Bengping mine is hidden over there . . . It’s the government that closed it. They put surveillance cameras all around the place.’ In May 2021, the Wall Street Journal reported that its journalist did manage to get very close to the mine on a mountain bike and saw that the entrance had become overgrown with vegetation. He was detained by the Chinese police for five hours and forced to delete a photograph of the mine from his mobile phone.

The Chinese government has shown much diligence and energy in keeping people away from the mine. Yet according to its own pronouncements, it has shown less interest in conducting further investigations at the Mojiang mine. Either it has been doing such work but keeping the results secret or it has decided not to look further into what can be learned from the site. Both possibilities are concerning.

So it is to satellite images that we turn. Today the site is blanketed with greenery but images obtained from April 2012 – when the miners who fell ill were at work – show a well-used dirt road leading to a set of buildings near the top of the hill and a well-trodden path lower down the slope to what may be the entrance of a horizontal tunnel, known as an ‘adit’ or ‘drift’. Based on research into similar sites by Brian Reed, an American engineer who has travelled extensively in rural China, including Yunnan, the buildings appear to be a U-shaped, prefab structure alongside storage tanks, typical of a mineral exploration camp. Two books unearthed by Reed reveal that the Bengping copper mining and smelting operation was originally one of the local, home-grown industrialisation ventures that were encouraged by Mao Zedong during the Great Leap Forward of the 1950s: ‘In 1958, Sun Zhongxiu, deputy secretary of the county party committee, organized nearly 10,000 people to mine the Bengping copper mine.’ But by 1960, owing to poor results, the mine had been abandoned. It appears there were several sites, one of which was referred to as ‘Bat Cave’, from which sixty tons of 6 per cent ore were extracted but not sold ‘due to inconvenient transportation’. Other records show that in 1978 an attempt was made to restart the mine. Then, according to official permits, sometime around 2011, at a time of high copper prices and a boom in copper exploration, someone seems to have had the idea of reopening the mine.

In early April 2012, we know that a small group of men began to clear bat droppings from the mine, which was full of the animals, some perhaps waking from hibernation, although in this subtropical area many stay active all year. The bats were of several different kinds, but especially numerous were the small, gregarious, insect-eating species known as horseshoe bats. It was warm, dusty and dirty work, inhaling the dust and noxious smell from the huge numbers of bats and their faeces.

It is not clear why the men were shovelling bat guano. According to the 2016 doctoral thesis, before the cases of the sickened miners, ‘many people had repeatedly entered and exited the abandoned mine, but no outbreak occurred’, which implies a regular trade was being carried out. It is probable that the miners had been contracted either to clear the mine for copper mining or to collect the guano to sell as organic fertiliser, or both. The bat guano trade is a lucrative one in some parts of the world, with small amounts of nitrogen- and phosphorus-rich guano being used in traditional Chinese medicine, as well as large amounts being supplied to farmers as fertiliser. It is collected from caves mainly in Mexico, Cuba, Jamaica, Madagascar, Indonesia and Thailand. A study of bat guano in a Thai cave, performed in 2006 and 2007 (but not published till 2013), found genetic material from coronaviruses in a small percentage of the samples and warned that bat guano miners should take preventative measures against exposure to dangerous viruses.

We do not know how many men worked in the Mojiang mine that April and May, but six were admitted to a hospital – not to the one in Pu’er City, less than a two-hour drive away, nor the one in Yuxi, less than a four-hour drive away, but to the one in the provincial capital of Kunming, almost a six-hour drive to the north. These six men had been referred by their local hospitals and clinics for specialised treatment at the Kunming Medical University Hospital.

A medical thesis

We know the story of the Mojiang miners only because of the medical thesis written by 李旭 (Li Xu), a student at Kunming Medical University. Completed in May 2013 and entitled ‘The Analysis of Six Patients with Severe Pneumonia Caused by Unknown Viruses’, the thesis concluded that the six miners had been infected by a SARS-related coronavirus (SARSrCoV) from bats, and the author noted that it was essential to investigate the bats in the mine.

This thesis had only been discovered and shared on Twitter in May 2020 by the anonymous user called the Seeker. One of us (Alina) translated critical parts of the thesis within a day and read, with increasing distress, about each of the patients and the struggle against the unknown disease with seemingly Covid-like symptoms. We separately visited the Chinese database of academic theses and confirmed that such a medical thesis existed. Attempts by us and others to reach the supervisor and the author to further authenticate the thesis were unsuccessful.

Dr Li’s 2013 medical thesis states that, after consulting Dr Zhong Nanshan, a serum immunoglobulin-M (IgM) antibody test was performed on the four living patients by the Wuhan Institute of Virology, to which the samples were sent. IgM antibodies are the body’s first line of defence in response to exposure to a pathogen, so a positive IgM test means that there has been a recent exposure. The WIV test results were indeed positive, suggesting a virus infection. The thesis did not specify exactly which type of virus the WIV had found IgM antibodies for. However, the 2016 doctoral thesis from the Chinese CDC deputy director’s laboratory (that had revealed the precise location of the mine) stated that the WIV had found the samples positive for SARS virus antibodies: ‘The blood test results of four patients showed that: four people carried SARS virus IgG antibodies, among which two of the discharged patients with higher antibody levels, and two hospitalised patients had lower antibody levels (Wuhan Institute of Virology).’ The two discharged patients were likely the two youngest miners, Li and Wu, who had left hospital at the end of May 2012, while the two surviving patients still in the hospital were likely Guo and Liu. Curiously, the doctoral thesis described the tests as finding IgG instead of IgM antibodies. IgG antibodies are produced during the initial infection but persist for months and sometimes years as a form of long-term protection in case the body encounters a similar pathogen again. Therefore, a positive IgG test could mean that the patients had been exposed to a SARS-like virus recently or perhaps months earlier. The terms ‘SARS-like’ and ‘SARS-related’ can be used interchangeably: both refer to coronaviruses of the genus betacoronavirus in the subgenus sarbecovirus.

According to Google Maps, the WIV is 1,885 kilometres from Bengpinghe by road by the fastest route: further than New York is from Orlando or as far as London is from Rome. Yet being the leading laboratory studying SARS-like and bat-borne viruses made the WIV an obvious choice to test the patient samples. These were sent to the WIV on Dr Zhong Nanshan’s instruction – to test for SARS antibodies. An immunoglobulin test result is not necessarily definitive proof of a virus having caused the disease, as opposed to some other pathogen, but both the 2013 medical thesis and the 2016 doctoral thesis certainly suggested a strong likelihood of the miners having sickened from a SARS-like virus.

Fortunately none of the patients seemed to have passed the virus on to healthcare workers or family members. The virus or viruses responsible for their illness very likely had not evolved to be as transmissible among humans as SARS-CoV-2 is. It had only managed to infect the six miners, perhaps, because they had been exposed to massive doses of the virus via the disturbed dust of bat guano within the confines of a poorly ventilated mineshaft for extended periods of time.

Virologists in the mine

No fewer than three teams of virologists visited the mine seeking the cause of the mysterious disease. According to their subsequent publications or theses, they each suspected that a virus had sickened the miners. Aside from the Chinese CDC deputy director’s group, another team came from the Beijing Institute of Pathogen Biology at the Chinese Academy of Medical Sciences and was led by Dr Jin Qi. They took samples from twenty bats, nine rats and five musk shrews captured in the mine. They identified a new Henipavirus-like paramyxovirus in three rats and reported this discovery in a 2014 article in Science magazine, titled ‘A New Killer Virus in China?’ The article was to be the first indication received by the outside world that there had been a lethal outbreak two years earlier. The rodent paramyxovirus they discovered was similar to ones carried by fruit bats that have caused occasional lethal outbreaks in humans in Australia and Bangladesh. However, this virus was incapable of replication in monkey, human or hamster cells in the laboratory. They concluded in 2014 that it was ‘more likely a curiosity’. While at the site, Dr Jin took a photograph that was later reproduced in Science magazine. It shows three scientists in full protective gear, gloves and masks handling specimens on a slope just outside the entrance to the mine. By lightening the background of the photograph, we were able to get a clear image of the entry to the adit, temporarily blocked by timbers and with what looked like more timbers supporting the tunnel’s ceiling and walls.

The WIV team was the most persistent. It was led by Dr Shi Zhengli. Between August 2012 and July 2013, Dr Shi’s group mounted at least four different expeditions to the mine, and at least three more in the subsequent two years. After all, this might prove to be the first ever recorded case of people catching a SARS-like virus directly from a bat. ‘The mineshaft stunk like hell,’ Dr Shi told a reporter in 2020. ‘Bat guano, covered in fungus, littered the cave.’

By 2013, Dr Shi’s lab had been searching caves in southern China for SARS-like viruses for close to a decade. Her group was part of the international consortium that had successfully tracked down a reservoir of SARS-like viruses – very similar to the one that had caused the 2002–3 outbreak – in a horseshoe bat of the species Rhinolophus sinicus, in a natural cave elsewhere in Yunnan province. Dr Shi’s group continued to sample thousands of bats in the wild, took swabs and blood samples from them, analysed the specimens for virus, sequenced the virus genomes, studied the interactions between viruses and cells in the laboratory, and even altered the genomes of those viruses in a bid to understand their biology and whether they could become human pathogens. She gained fame as the ‘Bat Woman’ of Chinese science even before the pandemic. By 2019, Dr Shi would be the deputy director of the WIV and director of its Center for Emerging Infectious Diseases.

After sampling hundreds of bats from the Mojiang mine, Dr Shi’s team discovered a single novel SARS-like coronavirus (SL-CoV) in 2013. They published a tiny part of its genomic sequence in 2016 under the name ‘BtCoV/4991’. It was found in a subtly different species of horseshoe bat from the one that by then was known to harbour the progenitors of SARS: Rhinolophus affinis, the intermediate horseshoe bat, rather than Rhinolophus sinicus, the Chinese rufous horseshoe bat. ‘We detected a SL-CoV-related sequence in R. affinis. This strain is distantly related to the previously discovered bat SL-CoVs in other Rhinolophus species and represents a new strain of SL-CoVs.’ Years later, in November 2020, Dr Shi revealed that at least another eight SARS-like coronaviruses had been discovered in the Mojiang mine by her team in 2015.

The lack of fanfare regarding these new sarbecoviruses was puzzling. Keeping in mind the mysterious SARS-like illness that had afflicted the workers in the mine, the BtCoV/4991 sample should have been of immense scientific interest. Here was a SARS-like virus, from a horseshoe bat, found at a site where three people had died from a SARS-like disease. It was exactly what the expeditions to the mine had been seeking. Yet Shi’s 2016 paper made no mention of the miners’ deaths. It was left to readers to join the dots by connecting this paper with Dr Jin Qi’s Science article about the Mojiang miners.

A sarbecovirus by any other name

Now the scene shifts to the city of Wuhan in central China in December 2019, when the Wuhan Center for Disease Control and Prevention detected a novel coronavirus in patients with atypical pneumonia. One of the first teams to obtain a sequence of the genome of the virus causing the disease was led by Dr Liu Yingle from Wuhan University’s State Key Laboratory of Virology. On 2 January 2020, they took samples from two unusual pneumonia patients in Zhongnan Hospital at Wuhan University: a thirty-nine-year-old man who worked at the Huanan seafood market in Wuhan and had fallen ill on 20 December 2019, and a twenty-one-year-old woman who had had contact with Huanan seafood market staff on 22 December. By 7 January, the scientists had sequenced the virus genome, something that would have been an extraordinary achievement twenty years earlier but is now fast and routine. On 8 January they looked in databases for a match and noticed that a section of the genome of the virus they had isolated shared a 98.7 per cent identity with BtCoV/4991, the published fragment of the genome of the virus found in the Mojiang mine in 2013 by the WIV. Dr Yingle’s paper was published in a prominent journal, Emerging Microbes & Infections, on 5 February 2020.

This ought to have been big news: a nearly 99 per cent match to the 4991 fragment was strikingly high and would have raised eyebrows, implying a possible connection to the outbreak. It got little attention, however, because two days earlier on 3 February a paper had been published in the prestigious journal Nature by Dr Shi’s team from the WIV just across town. In that paper, Dr Shi and her colleagues had also reported assembling a full genome sequence of the new virus. They found a 79.6 per cent match between the new virus and the 2002–3 SARS epidemic virus. Only after noting this did they then mention a match between part of the SARS-CoV-2 virus and ‘a short region’ of the genome of another bat coronavirus called RaTG13. This fragment, they said, derived from a virus sample collected from a Rhinolophus affinis bat from Yunnan province. They carried out full-length sequencing on this sample and found an overall 96.2 per cent genome match to SARS-CoV-2.

This discovery caused headlines around the world. It was the first evidence that the epidemic in Wuhan had been caused by a ‘new coronavirus of probable bat origin’, as the title of Dr Shi’s paper put it. At the time, there was speculation about the virus having come from snakes, bamboo rats or some other kind of creature sold in the Huanan seafood market – speculation encouraged by the Chinese authorities’ announcement on 22 January that the virus probably came from wild animals sold at the market.

One of us (Matt) recalls noticing with bafflement that the sentence announcing the discovery of RaTG13 had no citation: ‘We then found that a short region of RNA-dependent RNA polymerase (RdRp) from a bat coronavirus (BatCoV RaTG13) – which was previously detected in Rhinolophus affinis from Yunnan province – showed high sequence identity to 2019-nCoV.’ That was it: no footnote or reference. A day was wasted fruitlessly searching for the original report of the discovery of RaTG13. Where in Yunnan was it ‘previously detected’ and how? The internet and scientific literature had no mention of that name ‘RaTG13’ before 2020. Nor did the WIV paper provide a link to any previously published genetic sequence. Dr Shi’s group had, by early 2020, already described hundreds of bat coronaviruses from Yunnan province alone. None was called RaTG13.

What happened next was astonishing. On 20 February, Dr Vincent Racaniello, who is a virologist at Columbia University, published an article on his Virology Blog, speculating about how the new coronavirus had evolved naturally. In the comments beneath this essay, Dr Rossana Segreto of the University of Innsbruck in Austria noticed an anonymous post left on 1 March, saying there was a 98.7 per cent identity between part of the new virus and an entry in GenBank, the world’s public gene database.

Let’s meet Dr Segreto, because she is among the first of the volunteer investigators who played a vital role in disentangling the details of what happened – mostly in their spare time and at their own expense. Born and educated in Turin, it was there that Rossana Segreto earned a PhD studying the genetics of fungi that live symbiotically with orchids. She also met her German husband, a geo-ecologist, while they were both working on orchids. They moved to Norway, where she worked in Trondheim’s natural history museum, using genetics to sort out the relationships among species of mosses, corals and midges. Later, they moved to Germany, where Dr Segreto commuted across the Austrian border to work at the Institute of Microbiology at the University of Innsbruck. Here her work involved the genetic manipulation of Trichoderma fungi, ‘knocking out’ genes to understand their role in parasitising other fungi with the ultimate aim of using this fungus as a form of biological control in agriculture. Seeing how easy it would be to leak an organism from a laboratory by mistake, and with an increasingly deep knowledge of genetic manipulation, Dr Segreto was unpersuaded when, in early 2020, people started labelling the possibility that the novel coronavirus had leaked from a laboratory a conspiracy theory.

The comment Dr Segreto spotted on the blog on 1 March read ‘What about KP876546? 98.7% nt match.’ The serial number took her to the GenBank database, where a sequence had been deposited linked to the 2016 paper from Dr Shi’s group in which they announced the discovery of a virus with the name BtCoV/4991. The anonymous commenter was pointing out that one part of the SARS-CoV-2 genome was 98.7 per cent the same as a sequence from that bat virus. Dr Segreto downloaded the short sequence of 4991 and compared it with RaTG13. It was a 100 per cent match. She posted this revelation on the Virology Blog on 16 March and wrote to Nature to ask for clarification. She was told the authors would be contacted, but then nothing happened. So she continued to dig: ‘The more I discovered, the more I had to dig. It was possessive.’

She also came across the Wuhan University paper published on 5 February, which had identified the 98.7 per cent match between the new virus and 4991. Suddenly it all made sense. The WIV’s RaTG13 and 4991 were from the same sample from that mine in Mojiang, and these sequences both closely resembled the new coronavirus that was wreaking havoc in Wuhan and by then elsewhere. At this point, Dr Segreto had not yet joined the dots to find the outbreak among the miners.

On 7 March, the entry for 4991 in a Chinese database was altered to include a reference to RaTG13. This was later spotted by an anonymous Twitter user, schnufi666, who asked a close collaborator with the WIV, Dr Peter Daszak, the president of the New York-based EcoHealth Alliance, to confirm the identity of the two samples. On 9 May, Dr Daszak replied: ‘The answer is already in the papers & obvious to people working in virology’, which was curious given that it had been far from obvious to the Wuhan University team or Dr Segreto. It was not until July 2020 that Dr Shi would publicly state that RaTG13 and 4991 were the same thing. The Sunday Times in London ran a story, as did the BBC, which explicitly stated: ‘The BBC has had it confirmed, from the researchers running a respected Chinese database, that it is the same virus as one that the WIV previously referred to as RaBatCov/4991. There is no explanation for the change of name.’

A charitable interpretation is that the renaming may not have been intended to disguise the origin of the virus. Samples do get renamed occasionally and the original sequencing was of just a short stretch of the virus’s genome, whereas the new name now referred to a whole genome. Dr Daszak claimed in an interview that there had been a change in the naming conventions: ‘The conspiracy folks are saying there’s something suspicious about the change in name, but the world has changed in six years – the coding system has changed.’ (But, in 2021, it emerged that the virus that had been fully sequenced was still referred to as Ra4991 in a 2019 master’s thesis from the WIV.) Yet even if the naming system had changed, scientific readers would still expect an explicit reference to the prior publication of the sample in 2016. Why leave out the details of where it was found and the link to mysterious, unresolved cases of severe pneumonia? Certainly, the outcome of the renaming and lack of citation obfuscated the origin of RaTG13.

The real story of RaTG13

It was not until 18 May 2020 that the true story of RaTG13 was understood, thanks to the discovery of the 2013 Kunming medical thesis by the Seeker, an anonymous Twitter user – but no thanks to the WIV or Nature. The thesis had been dredged from cnki.net, the official website for master’s and PhD theses in China. On 20 May, the Seeker sent it to a husband-and-wife research team, Monali Rahalkar and Rahul Bahulikar, at the Agharkar Research Institute and the Central Research Station, Pune, India. They had just that day published a preprint online pointing out the electrifying similarity between 4991 and RaTG13. By 29 May, the Seeker had found and shared the 2016 doctoral thesis from the laboratory of the Chinese CDC deputy director.

In the pages that follow, we will partly rely on sources like the Seeker. Where the scientific and intelligence establishments had, in 2020, displayed only a surprisingly shallow interest (at least publicly) in the origin of the pandemic, these online sleuths have filled the gap. The Seeker’s real name is Prasenjit ‘Jeet’ Ray. He is a slim thirty-year-old, with shoulder-length hair, living in the city of Bhubaneswar in the Indian state of Odisha. With the exception of a brief excursion into Nepal, he has never left India. Many of those who have followed his repeated revelations and penetrating observations speculate that he might be a professional intelligence agent, or the spokesman for a team of data analysts. He has been accused of working for the CIA or Indian intelligence. The truth is more surprising and less dramatic. He is just a very clever young man who says he has prided himself on his internet research skills ever since his college days. ‘I learned how to make search engines work for me,’ he told us. ‘It was more madness than method.’ Born in Patna, he grew up in six different cities, as his father, a civil servant working in the airport and military sectors, moved around. As a child, he learned Sanskrit from a pundit who would cycle ten kilometres to teach him and he did well enough to debate in the ancient language. He went to college in Pune to study architecture, then took a certificate from the Zee Institute of Media Arts in Mumbai. Though he has worked as a science teacher and considers himself scientifically literate, he has no special expertise in medicine, biology or technology. But, time and again, we found that he combined a ruthless preference for facts over speculation with a sharp nose for new insights. Drawn to the question of where the virus came from, the Seeker posted his first findings in response to one of Alina’s tweets on 8 May 2020. Ten days later, rifling through the cnki.net website and using a login that he had been given on an open forum, he stumbled on the Kunming medical thesis, with its astounding tale of six men sickening, probably from a virus caught from bats and investigated by Wuhan virologists.

The Seeker became a member of a particularly tenacious, loose confederation on Twitter, calling itself the ‘Decentralized Radical Autonomous Search Team Investigating COVID-19’, or Drastic, that keeps turning up vital details about the possible source of the virus. Drastic describes itself as a ‘rag-tag team of internet researchers and sleuths’ who ‘came together on Twitter as a radical and subversive scientific collective to fight back against mainstream bamboozlement around the origins of COVID-19, and help spread the word about the dangers posed by unchecked gain-of-function research’. (We will come back to what ‘gain of function’ means in Chapter 8.) Drastic has no leader, but the closest it has to a coordinator goes by the pseudonym of Billy Bostickson, an indefatigable collator of information who claims to have a combination of experience with Chinese culture and expertise in biosafety and biosecurity. Prior to his involvement with Drastic, Bostickson told us that he had spent nearly a decade empowering ‘activists living under repressive regimes through collaborative sharing of electronic and audio-visual materials’. Drastic had its beginnings on Twitter, one of the few places that did not censor discussion of the virus origin throughout the first year and a half of the pandemic. Facebook flagged as ‘false information’ much of the work of these sleuths, even when it proved to be true information, while Reddit simply deleted it. Some of the two dozen or so people listed as members of Drastic seem more reliable, others less so; some are experts in biology and bioinformatics, but others are not; some use their real names while others are anonymous. They are not the sort of sources that authors usually turn to. But these internet sleuths have found nuggets of information that have since been confirmed more formally by qualified experts and even by the WIV itself. Bostickson came up with the name for Drastic: ‘We are decentralised and autonomous and a little radical.’

Reconstructing the events of early January 2020, when the SARS-CoV-2 virus genome was first sequenced, it must have been with a mixture of excitement and concern that Dr Shi found a near perfect match for the new virus genome from within her own laboratory: 4991 aka RaTG13. She later told Scientific American that she had frantically searched through her lab’s records to see if there was a possibility of the SARS-CoV-2 virus having leaked from the lab. Every laboratory-related incident was reviewed. None of the viruses her team had sampled from bat caves were a match to SARS-CoV-2. ‘That really took a load off my mind,’ said Dr Shi. ‘I had not slept a wink for days.’

Yet in publishing their discovery, Dr Shi and her colleagues neglected to connect the newly renamed RaTG13 to 4991, neglected to cite her own 2016 paper describing its discovery and origins, neglected to identify the mine where the bat sample had been collected, and neglected to mention that RaTG13 was from a site where three people had died of a respiratory illness of unexplained origin.

This was not the end of the revelations about how the seminal Nature paper by the WIV had been economical with the truth. The story of RaTG13 continued to evolve. Dr Peter Daszak of the EcoHealth Alliance, who had helped fund and had collaborated with Dr Shi’s group, gave an interview in which he asserted that the sample, whatever its name, had remained neglected in a freezer in Wuhan for six years until its similarity to the virus causing Covid-19 was noticed. He repeated this in a July interview with the Sunday Times. Only when the match was noticed in January had the entire sequence of that virus been assembled. That sequencing event, he claimed, depleted the sample so that nothing was left for further analysis. The WIV’s Nature paper had given the same impression. This turned out to be untrue. When the raw data underlying the RaTG13 genome sequence was uploaded to the GenBank database in May, its date stamps revealed that its various parts had been sequenced in 2017 and 2018. This discovery by internet sleuths and one of us (Alina) makes it all the more puzzling that the WIV scientists first fixated on the underwhelming 79.6 per cent similarity to a SARS virus, rather than the whopping 96.2 per cent similarity to a complete bat virus genome that was already in their database.

Once again, Dr Shi eventually confirmed this account of events deduced by sleuths. In her interview with Science magazine in July 2020, she conceded that her group had actually sequenced the complete genome of RaTG13 in 2018, not after the Covid-19 outbreak as originally implied by the Nature paper. But she failed to address the glaring question of why this had been obscured, in addition to the links to the Mojiang miners.

To summarise, an outbreak of mysterious pneumonia in a copper mine, more than 1,800 kilometres by road from Wuhan, led to patient samples being sent to Wuhan for analysis. A 2013 medical thesis concluded, after incorporating results shared by the WIV, that these miners had likely been infected by a SARS-like coronavirus from bats in the mine. An expedition by Wuhan virologists to seek the viral cause brought back hundreds of samples from bats. Their repeated visits to the mine turned up a bat-borne coronavirus in 2013, which was recognised to be a novel SARS-like coronavirus. The WIV team partly sequenced this new virus in 2017 and then fully sequenced it in 2018. When its sequence was found to closely match the sequence of the virus causing Covid-19, the Wuhan scientists published it under a new name and failed to cite their own paper detailing its discovery or to reveal that they had been studying the virus over the past few years or to mention that it had come from a mine where there had been a fatal outbreak of pneumonia.

The November addendum

Eventually, on 17 November 2020, Nature published an addendum to Dr Shi’s February paper. The addendum acknowledged the existence of the Kunming medical thesis; it confirmed the story about the Mojiang mine, the testing of the miners’ samples at the WIV, the new name assigned to the sequence from the 4991 sample, and the sequencing of its full genome by 2018 rather than 2020. It had taken nine months for the WIV scientists to come clean about the scintillating history of RaTG13. Dr Segreto, Dr Rahalkar, the Seeker and other sleuths had been correct about these points. They had been dismissed as conspiracy theorists at the time and derided by prominent virologists. Crucially, the addendum confirmed that ‘we suspected that the patients had been infected by an unknown virus’.

The addendum, however, contradicted both the 2013 medical and the 2016 doctoral theses by stating that ‘we tested the samples using PCR methods developed in our laboratory targeting the RNA-dependent RNA polymerases (RdRp) of Ebola virus, Nipah virus and bat SARSr-CoV Rp3, and all of the samples were negative for the presence of these viruses’. It also reported that they had tested for antibodies against the same three viruses with negative results. Remember that the medical thesis claimed that the WIV had detected IgM antibodies against viruses in the miners. The doctoral thesis said that the four living patients at the time of sampling carried IgG antibodies against SARS virus.

The addendum revealed two further startling pieces of information. Not only were the samples from four sick miners tested at the time for viruses, but they were retested after the Covid-19 pandemic began. In other words, the samples from the patients had been stored at the WIV throughout the intervening period, yet had not been shared with independent research groups outside of China for verification.

The second revelation was that far from turning up just one SARS-like, bat-borne virus in their expeditions to the Mojiang mine, the WIV team had in fact found at least nine: ‘Between 2012 and 2015, our group sampled bats once or twice a year in this cave [sic] and collected a total of 1,322 samples. From these samples, we detected 293 highly diverse coronaviruses, of which 284 were designated alphacoronaviruses and 9 were designated betacoronaviruses on the basis of partial RdRp sequences. All of the nine betacoronaviruses are SARSr-CoVs.’ The WIV did not disclose more information about these other eight SARSrCoVs.

The mine in Mojiang County remains, in 2021, one of the closest clues yet found to the origin of the pandemic: it is the place where the virus most closely matching SARS-CoV-2 was found – a virus that was studied in a Wuhan laboratory before the detection of the Covid-19 outbreak.

Four months after the Nature addendum, the WIV appeared to revise its story; it told members of the WHO-China global study of origins of SARS-CoV-2 that ‘the reported illnesses associated with the miners, according to the WIV experts, were more likely explained by fungal infections acquired when removing a thick layer of guano’.