It Came From A Lab!

It’s being widely reported SARS-CoV-2 might have infected humans via dogs.

In China many already believe that pets spread the new coronavirus which has led to – amongst other things – dogs and cats being thrown from apartment balconies and killed but this study proposes a mechanism. A mechanism I find entirely unconvincing.

We’ve been looking for the intermediate animal host between bats and humans since SARS-CoV-2 first appeared and many animals from snakes to pangolins have been suggested. A problem is the viruses isolated from them are quite divergent from SARS-CoV-2 – which suggests a common ancestor further back in time. So is this the answer?

First, this is not a ‘real life’ study. It’s a computer model based on genetics. Just like the ‘spread by runners’ or ‘spread in supermarkets’ stories it has no basis in hard facts. It provides no new data. It’s just conjecture based on a computer model.

They do manage to weave a new hypothesis re how SARS-CoV-2 might have evolved to infect humans that’s rich in inference but has absolutely nothing concrete it that supports the notion that the dog did it. I’ll come to why it entirely lacks plausibility in a moment – just let me briefly unpack it for you first.

What they did was compare the molecular signatures of viruses in many different hosts and looked at changes and adaptations you can find within viral genetic material that arise as they fight off and evade a new host’s immune system. There is a constant war between us and viruses and you can see the scars in both genomes. Which is pretty cool.

The study suggests – based on genetics – that the intermediate host between bats and us might be dogs or some other canid.

Mammals – including us – have an antiviral protein, aptly called ZAP (zinc finger antiviral protein), which attacks viral RNA. It looks for certain sequences in the viral four-letter genetic alphabet (called CpG dinucleotides – a cytosine is followed by a guanine), binds to them and marks them for destruction.

But because of a thing called evolution, viruses have got wise to that. Viruses that have fewer CpG dinucleotides can resist ZAP because they have fewer places it can attack. So evolution is kinder to them and they’re more likely to reproduce. HIV also exploits this evolutionary trick to fool human antiviral defences, incidentally.

By comparing many different coronaviruses, the study authors found that only genomes from canine coronaviruses (CCoVs) had CpG values similar to those observed in SARS-CoV-2 and BatCoV RaTG13. They also found canids have coronaviruses infecting their digestive tract with CpG values lower than those that infect their respiratory systems. Which is interesting to nerds like me. Camels have CoVs in their digestive tracts too – remember MERS?

What the authors then do is suggest that because dogs spend so much time licking their own and other dogs’ arseholes, that is a potential mechanism for a gastrointestinal virus to get into the lungs. Which again is an interesting idea – but it’s a hell of a stretch. In fact it reeks of trying to graft relevance onto something a computer model puked out which is pure artefact.

Yes, SARS-CoV-2 may have infected humans from a ZAP-rich intermediate or evolved in a tissue with high ZAP expression. But many viruses have lower CpG dinucleotide levels than SARS-CoV-2 and while this is a useful viral trick to evade mammalian defences, low CpG levels does not necessarily a serious pathogen make.

But the real challenge here is twofold.

The first challenge is, given that SARS-CoV-2 must have been lurking in the background for quite some time and is highly infectious, we can infer that its transfer from Whatever-The-Fuck-It-Was to human is a very rare event. Consensus is it happened only once.

Were that not the case there would have been multiple outbreaks instead of a single, localised origin.

Again, applying logic one or more of these statements MUST apply:

None of which really apply to Fido, do they?

The second challenge is rather easier to get your head around.

It’s been shown SARS-CoV-2 really doesn’t infect dogs very well.

In fact, it’s really hard to infect dogs with it, full stop. Dogs have coronaviruses, sure. But dog coronaviruses and human coronaviruses are different. There’s no evidence of cross-infection.

Yes, you could conceivably construct a pathway by which this might happen under certain circumstances but it’s really, really unlikely. Were this a thing we’d know about it by now. It replicates poorly in pigs and chickens too.

So NO, transmission between Fido and us really isn’t something we need worry about.

But SARS-CoV-2 does replicate efficiently in ferrets. And cats. I knew the Fluffy Vermin had a paw in this somewhere. Bastards.