13 July 2020

How do we know COVID isn’t a bioweapon?


It’s a virus that has infected more than 12 million people and caused more than half a million deaths worldwide.

If you have listened to US President Donald Trump and several of his colleagues, the leader of Iran, and official sources in China at various points, this threat came not from nature but was intentional.

The conspiracy theory that SARS-CoV-2 actually came from a laboratory has gained traction in recent months, with many accounts blaming the pandemic on a leak (or malicious release) from the Wuhan Institute of Virology in China near where the virus was first detected.

Whether the politicians themselves believe this conspiracy, or it is simply a political ploy, the sticking power has prompted many in the scientific community to try to debunk such claims.

But what makes the scientists so certain? When gene-editing technology is progressing at such a rapid rate, and there are still major questions around the origin of the virus, how can we really know that we aren’t victims of a bioterrorist attack?

We’ll probably never be 100% certain, says virologist Dr Megan Steain, lecturer in infectious diseases & immunology at the University of Sydney. “But I’d say we’re pretty confident – 95-to-98% confident – that this is naturally occurring,” she says.

This is because the genome has been sequenced by many labs around the world and pored over by scientists.

Like archaeologists looking at fossils for clues about a creature’s evolution, virologists analyse the genetic code for signs of human interference. They have compared the sequence to those from other similar viruses to find out what it’s most closely related to. This is used to create what’s called a phylogenetic tree, which looks just like an ancestry tree, to show us the virus’s closest relatives.

So far, the closest relative scientists have found to SARS-CoV-2 is one found in horseshoe bats known as Bat CoV RaTG13. Importantly, the major difference in the coronavirus affecting humans, and the part that makes it so effective at infecting humans, is its spike protein.

Dr Christian Stevens, at the Mount Sinai university in the US, explained in his blog why the receptor binding domain on the spike protein made it such a threat.

Firstly, it has tweaks that make it far better at binding to ACE2 receptors on our cells than other coronaviruses, such as the first SARS virus. It is also able to work on a range of cells and tissues in our bodies.

And lastly, it can be coated in sugar molecules, known as a glycan shield, which appears to help it hide from our body’s immune systems, he says.

Conspiracy theorists believe that this perfect set of tools must be the result of human engineering.

However, creating this virus in a lab and knowing that it could be a pandemic-inducing pathogen seems to be more in the realms of science fiction than present-day science.

If you were a scientist designing such a virus, the first thing you would need is an existing template, explains Dr Steain. “You don’t just assemble it from nothing in the lab.”

We would expect engineers to choose a virus that already causes disease, such as the first SARS-coronavirus, as the backbone for the new virus, she says.

“And if you were to do that, you can’t make sweeping changes across the whole genome, you would generally just mutate in specific regions as you go.”

The end result would be something that looks a lot like the virus backbone that you started with, but has particular mutations in a particular region.

“Whereas when we look at the genome to this virus, it doesn’t look like it’s come from any of those existing backbones,” Dr Steain says.

Instead, the virus with the most genetic similarities, RaTG13, is unlikely to be able to efficiently infect human cells. This would make it an unlikely candidate to base your weapon on if you were a scientist aiming to make a deadly and very infectious virus.

Others have jumped on a preprint that claimed to have found HIV-like insertions on the coronavirus genome, suggesting the virus may have been engineered in a lab. The authors wound up retracting the paper when they realised that these insertions were much more similar to mammals, insects, bacteria and other viruses, than HIV-1.

There’s another scenario though, in which scientists took a coronavirus and put it in lab conditions that sped up its evolution. This is known as simulated natural selection.

This would create a virus without the obvious signs of human interference.

However, Dr Steain thinks this, too, is improbable.

For starters, the sugar molecules that help the virus avoid detection by antibodies only evolve when a virus is trying to hide from an immune system.

“When you’re culturing the virus in the lab in cells, there’s no immune system pressure there,” says Dr Steain. “So there’s no reason that the virus would evolve over time to gain that mutation.”

And the features that make this virus so good at infecting humans are things that scientists wouldn’t at all predict, either.

Computer models can help predict how mutations could affect the function of the virus. However, they aren’t accurate and it would be difficult to predict what mutation worked best in reality.

In particular, models suggest the spike protein on SARS-CoV-2 wouldn’t actually bind that well to the ACE2 receptor, giving scientists no reason to engineer a bioweapon with it.“By mutating something you might gain a function, but you might lose another function at the same time,” Dr Steain says.

Moreover, the virus doesn’t appear to have the telltale signs you would expect if it was created through rapid evolution in a lab. Each time a virus replicates, random errors are introduced. Some of these mutations will actually change how the virus functions, and others won’t do anything.

Mutations that don’t change anything should occur pretty consistently, because they have no effect on the virus’s survival. Mutations that do change the virus’s function, on the other hand, will happen more or less commonly depending on how well they suit the environment the virus lives in.

As Dr Stevens explains in his blog, if the virus was being forced to rapidly evolve in a laboratory, you would expect to see plenty of changes that affect the virus’s function in at least some part of the virus’ genome.

However, an analysis by virologist Associate Professor Trevor Bedford, at the US Fred Hutch research centre, indicates the ratio of mutations found in this novel coronavirus is about the same as you would expect from a virus that arose out of natural selection, not accelerated evolution like that found in a lab.

So the steps that the bioengineers would have needed to follow in order intentionally to create this virus would have been illogical from a virologist’s perspective.

“Nobody in the world could have predicted that the changes that had occurred naturally would have caused this virus to be so pathogenic,” says Professor Andreas Suhrbier, head of the Inflammation Biology Laboratory at the QIMR Berghofer Medical Research Institute.

It takes scientists many years to understand what effect certain genetic changes have on the way the virus functions in the community, so there was no rational basis for choosing a bat virus or choosing to re-engineer the spike protein.

“If you’re trying to make a biological weapon, why would you re-engineer a harmless virus?” Professor Suhrbier asks.

Instead, malicious actors have a host of known pathogenic viruses and bacteria that they could, and in the past have, modified to turn them into effective warfare agents.

Ultimately, we still don’t know the exact origins of SARS-CoV-2. This means that questions may remain until a smoking gun is found.

Some believe that this will be finding a wild animal infected with a virus that looks extremely similar to this virus, probably in the region close to the first outbreak.

But with most resources currently diverted to combating the spread of the virus, and treating its victims, we may be waiting for some years.


This article was originally published under the headline ‘How we know COVID didn’t come from a lab’.

COVID-19 live update
Something to say?

Leave a Reply

8 Comments on "How do we know COVID isn’t a bioweapon?"

Please log in in to leave a comment

Sort by:   newest | oldest | most voted
Jake Lutz
Jake Lutz
1 month 2 days ago

Best article on the subject yet, but still leaves me wondering if this was weaponized. Having enough test subjects, decades of time, government control, and a very quiet place out of the eyes of the world. I think it’s possible using trial and error plus control, no labs needed just stiff protocols and no morals.

Dr. I Rivlin
Dr. I Rivlin
3 months 18 days ago

What difference does it make if it’s a bioweapon or not? It’s provenance (or creators’ malicious intent) doesn’t affect the infectivity or social impact.
We should be coming up with a rational, appropriate and affordable plan, that secures our young people’s future.
All I can see, ATM, are tyrannical politicians trying to outdo each other’s (faux) compassion and re-electability.
We’re witnessing the most bizarre social upheaval imaginable.

Alan Jules Weberman
8 months 22 days ago

Proximity to WIV, not found in animals, unique polybasic furin cleav. site. Not a bio weapon but a lab leak.

8 months 27 days ago

Is everything the virus does , a passive movement ,and just attaches itself to receptives on our cells.
If not a passive attachment,, how (without a brain), does it decide where to go, when and how to mutate, when to hibernate , and when to reappear

Joe Kosterich
8 months 27 days ago

As the article states- we don’t know. Incompetence is a better explanation than malevolence. It is most likely that the virus escaped the lab by mistake. It was not a deliberate ploy!

Dr Angela Carter
Dr Angela Carter
8 months 27 days ago

Completely agree Joe. There is a very distinct difference between “came from a lab” which is in the title, to “created in a lab”, which is all that the article discusses/attempts to refute. Either the person writing this article hasn’t thought this through carefully, or there is a deliberate attempt to obfuscate. Or most likely, a bit of both.