Critique of my Colleagues
The molecular evidence of a man made origin should not be dismissed
I work closely with people who I disagree with on certain topics. This is not only healthy, it is important. We are not part of a cult.
Prof Martin Neil (MN) and Dr Jonathan Engler (JE) have just published a detailed argument setting out how claims of gain of function (or as they call it “claim of function”) are detracting from the more important argument about the extent of policy induced harm.
Agreement
Their fundamental point that the impact of any virus must be judged on the overall data of the impact it actually had in the real world is absolutely right and very important. No matter how bad something looks on paper, it is only as bad as it is in the real world. What might have a certain impact on cells in a laboratory - or even in non-human species, especially rodents - is not guaranteed to have a similar impact in humans during an infection.
I agree with their thorough analysis showing that most attempts at “gain of function” result in loss of function. I too do not believe the threat from man made viruses to be anything like what is claimed. Nature has already created a full spectrum of pathogens. All that man made mimics can do is fall somewhere in that spectrum. The fact that SARS-CoV-2 was no match for human immune systems, despite how nasty it looked on paper, is evidence to that effect.
The threat from “gain of function” often comes down to a classic fear of the enhancement of a hemorrhagic virus spread by bodily fluids, like Ebola, to be more transmissible. Ebola and coronaviruses are very different viruses, with distinct transmission mechanisms, structures, ways of attacking the body, and methods of evading immunity. Trying to combine them in a lab would be akin to attempting to create a cross between a bear and a wolf and expecting to produce a wolf-sized bear that poses a greater threat. Even a bear-sized wolf, while it might sound scarier, would collapse on its spindly legs and be immobile. The fundamental differences in their biology make such a hybrid impossible to create and highly unlikely to function.
I also agree that virus hunting, in which more and more resources are being invested, is a sure way to find “novel” viruses, mostly of questionable relevance. The current definition of a “pandemic” as being the spread of any “novel virus” regardless of its impact, means pandemics will be declared in situations where the kind of (albeit minimal: “catch it, kill it, bin it”) response suggested in pandemic plans would be utterly inappropriate as it would cause unnecessary alarm.
MN and JE quote the Coronavirus Study Group (CSG) International Committee on Taxonomy of Viruses from February 2020, who said, “If the strict match criterion of novelty was to be applied to RNA viruses, it would have qualified every virus with a sequenced genome as a novel virus.” They went on to decide that SARS-CoV-2 was not novel compared to SARS1 but this section was removed from the peer reviewed article published just a few weeks later in Nature. SARS-CoV-2 was not considered genetically distinct from the broad family of SARS-like coronaviruses (of which there are many) even though there were clinical differences. I can agree when MN and JE say, “Together the assignment of the word novel and the use of the name looks like the result of contentious political process or social construct, presented as a scientific one where all uncertainty, doubt and dispute have been stripped out.” However, they argue that the fact it was so similar to existing viruses meant it was not noteworthy. I would argue that the semantics of virus taxonomy is not as important as the clinical impact. While I do not think the clinical differences were as marked as some claim, they existed.
There are other points on which we disagree.
Disagreement
I think the evidence of a man made origin, based on what is seen in the genetic sequence, is almost watertight. Even the viral taxonomy committee said in their paper that the introduction of both SARS-1 and SARS-CoV-2 “was likely facilitated by independent unknown external factors.” There are too many unique oddities about this sequence to dismiss it as from just another member of the coronavirus family. At times it seems MN and JE’s proposals for explaining away the evidence of gain of function feel like clutching at straws with each argument excused with separate reasoning.
Furin Cleavage Site
MN and JE claim the furin cleavage site was common among closely related coronaviruses. They are right but the most closely related virus to have this sequence is only distantly related with only 36 percent of the sequence the same as SARS-CoV-2. However, what is particularly suspicious about it is that the positioning of these extra letters in the genetic code is identical to the position used in engineered coronaviruses to make them more infective and more lethal.
Furthermore, it takes three letters to code for each amino acid, but there is more than one way to code for each. There are six different three letter combinations all of which code for the amino acid arginine but SARS viruses do not favour CGG, only using CGG 1.5 percent of the time. Yet, in the furin cleavage site there are two sequential CGG codes. There is no other point in any of the SARS-like genomes where two CGGs occur next to each other. CGG is a foreign word as far as SARS coding goes. In contrast, CGG is common in human sequences and a favourite for laboratories when coding for arginine.
HIV sequences
In January 2020, Prashant Pradhan and colleagues identified four unique regions in the spike protein that did not match any other coronaviruses in international databases. While these databases do not claim to contain more than a sample of viruses, they do represent a cross section indicating the range of coronavirus sequences. They noticed that all four matched with sequences found in HIV (although some were rare HIV subtypes). One had 4 amino acids which matched the Furin Cleavage Site and a separate run of four needed for its function. Three specifically matching parts of the gp120 protein, which enables HIV to enter immune cells. An argument could be made that each individual part of the sequence of 6 to 12 amino acids could have been incorporated at random - as coronaviruses do “borrow” sequences from other viruses. However, three of the HIV sequences, while quite separate in the sequence, code for a total of 24 amino acids that are physically aligned once the protein is folded. They are positioned right at the receptor binding domain which is precisely where they would need to be to be functional. The chances of that happening by chance are astronomically low.
The paper was withdrawn, after two days on a Sunday, but the authors continue to allow access to it through their researchgate page and commented the withdrawal was due to “pressure from the people with vested interests.”
Furthermore, these inserts are really neat (as in tidy). When coronaviruses steal RNA from other viruses it happens at random. While some of it might prove functional, there is almost always some hanger-on sequence that is of no relevance. The inserts that were present in SARS-CoV-2 were edited right to the edge of just the functional parts.
MN and JE fail to explain these points beyond saying, “Even the gp120 HIV inserts might have originated from wherever that sequence can be found elsewhere in nature and in man.”
Sites matched for enzymes
When manipulating genetic material in a laboratory, certain enzymes are used to cut and glue segments of sequence to assemble what is needed. In years gone by, genetic manipulation would leave a trace in the code where it had been snipped apart and then metaphorically glued back together again. But Ralph Baric described in 2002 how to edit viruses using what he called a “no see’m” approach, which meant the sequences would appear to have arisen naturally.
In nature, the sites at which these enzymes cut are randomly distributed through the sequence but in the SARS-CoV-2 virus they are very evenly distributed. Efficient laboratory manipulation of viruses starts with ensuring even distribution of these cutting sites. The ordered pattern that was present screams a synthetic origin. Even more damning, the location of two of these sites was identical to sites published after laboratory manipulation of a bat coronavirus by the WIV in 2017.
MN and JE quote a paper written by Wu (who graduated from the WIV in 2012). This paper makes the claim that the fact the distribution was even at the moment in time when SARS-CoV-2 was first sequences was down to chance alone. It was as if the random pattern, which does change with time, stumbled upon a perfect distribution which has since become random again. They did show that among thousands of samples it is possible to occasionally find one that is evenly distributed by chance. However, finding it in a virus with other signs of genetic manipulation would be a massive coincidence.
A further point is made regarding the cost of engineering such a number of evenly spaced restriction sites which was claimed to be prohibitive at $10 billion. However, the calculations were carried out assuming only old fashioned techniques were used. Using genetic engineering techniques the project would come to a few million dollars.
It is far too much of a stretch to claim these could have ALL been natural coincidences.
In fact, there are other clear indicators of gain of function research. Ralph Baric, a virologist based in North Carolina, had written a paper about how mutating a particular gene (nsp14) could lead to a 21 fold increase in mutation rate in coronaviruses. The replication rate of SARS-CoV-2 is faster than natural coronaviruses, thanks to a particular mutation in that very gene. Moreover, the DEFUSE grant proposal from Peter Daszak to the NIH mentioned nearly all of the anomalies. The funding was denied but scientists often carry out work that has been included in such proposals either before or after they are submitted and these researchers had plenty of funding from elsewhere.
The Department of Defence funding went instead to a very similar proposal from another group called PREEMPT. Dr Danielle Anderson was sent to the WIV to do the work but left in November 2019. Her supervisor at Duke University, Linfa Wang, sent an email predicting the virus would be “another WMD scandal” on the 5th January, when there were only 44 cases of pneumonia. Some claim he resigned from his role as director that he held for a decade on 10th January, the day the genome was made public.
There are actually four laboratories in Wuhan. The other three were a Chinese CDC lab, which worked on bat coronaviruses and had a lower safety authorisation than WIV, a military lab and a German-Chinese viral laboratory in Union hospital. The fact these never get mentioned is a story in itself. The Union hospital laboratory was the base for a German-Chinese virology collaboration, which included the WIV and Tongji Medical College and had already received over a decade of government funding. The research aimed to create vaccines and treatments for chronic viruses including HIV. The lead German virologist was Ulf Dittmer who has links to Christian Drosten. Christian Drosten had apparently started work on a PCR test in 2019 a week before the Chinese had said the pneumonia had been caused by a coronavirus and published inaccurately on asymptomatic spread in January 2020. He worked in collaboration with scientists at WIV and his hospital has close contacts with Tongji hospital.
Ultimately, I cannot tell you who made it based on this evidence, only that it was man made.
Molecular Clock
Looking at how mutations developed and tracing back to a common ancestor with molecular clock analysis shows a common ancestor for SARS-CoV-2 was from mid-September to early October 2019. This fits with plenty of other evidence about the earliest cases based on antibodies in stored blood samples and virus in wastewater. MN and JE dismiss these findings saying they could have been cross reacting with other coronaviruses without explaining how there was no positivity before October 2019.
My different interpretation rests largely on my having a different hypothesis regarding viral surges. MN and JE’s position is that any viral spread from a point source must result in a surge with the same timing globally. I dispute that. There is an evidence base of many decades, if not hundreds of years, showing that geographical, seasonal, immune and other factors led to surges of influenza in a manner that was not entirely predictable and not entirely understood. Such a “seasonal trigger” would account for minimal impact in Autumn 2019 and 2020 while a viral surge occurred in spring 2020 in Western Europe and elsewhere. In the meantime, a lack of spring surge in Eastern Europe (including Germany), South East Asia and Australia kept them covid free. Ultimately such surges have been seen everywhere with a pattern reminiscent of influenza in those places.
MN and JE point out that challenge studies have failed and that any claim of function cannot therefore be validated. I would argue this is because only a fraction of the population is susceptible to any one variant. Once that variant has passed through a population, the population is made up of those who were never susceptible and those who have since acquired immunity.
The authors argue that gain of function research is unethical as it presents only risk with no benefit. I am less eager to see a total ban on virological research in general, much of which could be caught up in a gain of function research ban.
The more fundamental point that we agree on is that a virus was not the sole cause of the death spikes. The accumulating evidence of dystopian policies, which in certain cases amounted to euthanasia, has somewhat taken me aback and all such evidence needs to be accounted for when estimating the impact of the virus alone.
Policy induced harm
SARS-CoV-2 had an apparent mortality in line with a bad seasonal influenza when overall mortality is compared per 100,000 people.
Figure 1: Weekly mortality per 100,000 from 1983-2022
However, that mortality rate takes no account of a variety of iatrogenic harm that will have inflated the excess mortality. There were multiple factors that led to a higher mortality in both covid and non-covid patients.
This includes:
Blanket Do Not Resuscitate orders were issued at a time when hospitals were refusing admission to anyone with a DNR.
Normal treatment of community acquired influenza ceased and even hospital patients were not given antibiotics they otherwise would have had.
People were aggressively ventilated after advice that not doing so would expose staff to aerosols. One doctor commented, “Never in my life have I had to ask a patient to get off the telephone because it was time to put in a breathing tube.” Ventilated patients had a higher mortality rate.
Patients who deteriorated were started on respiratory depressant drugs that would have accelerated death. Palliative care experts called out this problem in May 2020.
Care home residents had vital support from loved ones removed at a time when staffing was low and staff believed that they might die if they didn’t use PPE that was in short supply. The consequence was inevitably that the frail did not receive the regular attention that they had before and which saves lives not least by keeping them hydrated.
People were deliberately terrified and fear has a known detrimental effect on immunity.
Worse, from February 2020, people were scared to attend hospitals when they needed care. Even operations for life threatening conditions fell dramatically. Some were denied care because they had a cough or fever, even though they were dying and indeed did die.
I do concede that there were more intensive care stays with covid than influenza and they were for longer periods. In total there were 10,000 such intensive care stays in England in Spring 2020. The NHS could probably do with having more intensive care beds and covid exposed that failing. For decades the NHS business model has been to cancel elective surgery that requires intensive care admission in order to free up beds for patients with respiratory viruses. This business model was always risky given the known variation in illness from respiratory viruses year to year.
Ultimately, we all agree that there was significant iatrogenic harm and that understanding the harm from covid can only be achieved after subtracting this unnecessary harm. The big question being, had there been no news coverage at all in this period, no change in protocols, no testing, what would the all cause mortality have been? In such a world, would any rise in mortality that did occur have been thought newsworthy in retrospect? I think not.
Finally, I do not think we should allow the debate over any nasty (on paper) sequences in a virus to detract from the important points about iatrogenic harm. That is the main story because that, unlike the spread of an airborne virus, was entirely preventable. It needs to be acknowledged if we want to prevent future harm. There is every reason to fear that society is walking towards a repeat. Systems are all aligned for the detection of the next “novel” virus, hyping up fear, producing panicked protocols and stopping normal treatment of patients (all while governments tighten the screws and increase their power). If another gain of function virus is detected, or indeed a zoonotic one, the key to minimising loss of life will be in remaining calm and rational and not being manipulated by fear propaganda.
I will end with the same last sentence as they used “If there actually was a function gained by the virus it was the power to help trick humanity into a dramatic act of self-harm.”
Excellent article, Clare. Thank you for this.
Oh, how I admire you for this article, Clare!
Measured, fair, civilized critique.
Unlike some in this movement- viciously attacking others for differing views.
Thank you.