I don’t understand your argument here. We’re looking at SARS-CoV-2 kicking around Wuhan for some time before December regardless of how it got there. I don’t see how this has any bearing on what the source was. I don’t think Bloom is asserting that earlier versions of the virus were any less competent at spreading in humans.
1 Like
It means that if it was some bug released/created that was super effective at infecting humans, then it should confuse you that it wasn’t super effective at infecting humans for months.
I’ve read the whole Nature paper now, they are basically saying that their data is consistent with two theories: a) a gain-of-function experiment gone wrong or b) pangolins as an intermediate host between bats and humans, involving a recombination event. While they’re clear that option B is a realistic possibility, there are reasons to doubt this happened, laid out in the section “Implications for the original SARS-CoV-2 animal source”, which I’ll reproduce here as I think it’s interesting:
A coronavirus isolated from Malayan pangolins (pangolin-CoV) shared the same S protein RDB sequence as SARS-CoV-2, raising the suggestion that pangolins may have acted as an intermediate SARS-CoV-2 host between bats and humans. Although it has high spike RBD sequence similarity to SARS-CoV-2, pangolin-CoVs are not closely related to SARS-CoV-2, with ~ 90% sequence similarity across their whole genome23. It is noteworthy that the RBD common to both pangolin CoV and SARS-CoV-2 binds strongly to both pangolin and human ACE2, despite significant differences in these ACE2 molecules with only 63% of their binding site residues being common (Table 1), and the sequence similarity of ACE2 is only marginally higher between pangolins and human ACE2 ~ 85%, than between and bat and human ACE2 ~ 82%. Remarkably, Pangolin-CoV S protein has 100% amino acid identity with SARS-CoV-2 S protein but has much lower levels of identity of 98.6, 97.8 and 90.7% in the E, M and N proteins24. Pangolin CoVs isolated from Malayan pangolins from two different regions in China showed differences in the residues interacting with human ACE223. One possibility might be that a pangolin was simultaneously co-infected with a bat ancestor to SARS-CoV-2 at the same time as being infected by a pangolin CoV. This could have allowed a recombination event to occur whereby the spike RBD of the pangolin CoV was inserted into the S protein of the bat CoV, thereby conferring the bat CoV with high binding for both pangolin and human ACE2. Such recombination events occur with other RNA viruses and explain creation of some pandemic influenza strains56. However, such events are rare as they require coinfection of the one host with two viruses at exactly the same time and the SARS-CoV-2 genome was reported to exhibit no evidence of recent recombination, arguing against this possibility57. Most importantly, if such a recombination event had occurred in pangolins it would be expected to have triggered an epidemic spread of the new highly permissive SARS-CoV-2-like virus among pangolin populations. Currently there is no evidence of a pangolin SARS-CoV-2 outbreak, making this scenario unlikely. Indeed, pangolins might be predicted to be protected from SARS-CoV-2 infection by the existence of cross-protective neutralising antibodies against pangolin coronaviruses given their close RBD similarity, making it even less likely that a SARS-CoV-2 was widely infecting pangolin populations and indeed no evidence of any such infection has been reported. Notably, all pangolin coronaviruses identified to date lack the furin-like cleavage site between S1/S2 in the SARS-CoV-2 S protein that facilitates its rapid spread through human populations. The fact pangolin CoV S protein does not have the furin cleavage site that is a prominent feature of the SARS-CoV-2 S protein45, argues against pangolins being the intermediate vector for transmission of SARS-CoV-2 to humans. The major similarity of SARS-CoV-2 to pangolin-CoV lies in the S protein RBD residues that SARS-CoV-2 acquired by some unknown mechanism.
To be clear, I’m not taking this paper as gospel, perhaps they’re a little too certain in their conclusions, but it’s certainly interesting.
They’re critical of earlier assessments of binding affinities, which don’t accord well with empirical assessment of infection susceptibility in diverse species. It’s in the paper under the section “Comparisons of our model predictions to other studies”.
What the Nature paper is saying is that the virus is, in fact, super effective at infecting humans, that’s a fact regardless of how you think it got into the human population. Assuming a zoonotic origin doesn’t help explain why there weren’t large-scale outbreaks earlier. I don’t think Bloom is positing that the earlier iterations were any less effective at infecting humans, they were substantially similar to the genotypes found in the wet market outbreak.
We’ve seen repeatedly that it’s possible for the virus to bubble under the surface for some time before some super-spreader event launches it into detection range, this happened in Italy in the early stages of the pandemic.
This is not disputed.
Again, if something was released that was super infective from a lab, you would expect it to be super infective right away. I don’t think you understand how this applies to your discussion, but I can’t think of another way to explain it.
My reading of all this could be wrong, but If Bloom is correct and if it originated from either the market or the Wuhan lab, you have the rather curious situation that it somehow kicked around a massive city for 2+ months before blowing up at the Wuhan market in December and getting everyone’s attention. That’s especially odd if this was a secret engineered bioweapon.
A better fit is that this circulated at low levels in outlying areas outside Wuhan for a while, someone took a trip to the city, and both the wet market and biolab are largely circumstantial.
Where are you getting 2+ months from? Wikipedia on the origins of the pandemic says this:
Based on retrospective analysis published in The Lancet in late January, the first confirmed patient started experiencing symptoms on 1 December 2019, though the South China Morning Post later reported that a retrospective analysis showed the first case may have been a 55-year-old patient from Hubei province as early as 17 November.
The Huanan outbreak was noticed on December 26, so we’re talking about 6 weeks. I basically don’t think that a gap of 6 weeks from patient zero to the discovery of an outbreak is evidence of anything. I think that sounds plausible. I also still don’t see how Bloom’s data has any bearing on this, we already knew that the virus was present in Wuhan some time before the Huanan outbreak. Your theory about circulating outside Wuhan is speculation and is entirely beside the point, since we know for a fact that the virus was present in Wuhan on December 1 at the absolute latest. That the virus was able to go at least 25 days from patient zero to outbreak detection while in a major city, while already at full competence in terms of human transmission, is a fact regardless of anything else.
The main point of Bloom’s reconstruction is that it shows the virus must have been circulating way earlier than December. He mentions a paper by Kumar, which puts it as Sep-Oct. (I was bored last night, I kinda took a dive into the weeds on this)
https://twitter.com/jbloom_lab/status/1407445628049756164?s=20
https://twitter.com/jbloom_lab/status/1407445629496823809?s=20
https://twitter.com/jbloom_lab/status/1407445630612480001?s=20
OK. Well that seems kind of speculative, but from that paper:
Because proCoV2 is three bases different from the Wuhan-1 genome, we estimate that the divergence of the earliest variants of proCoV2 occurred 5.8–8.1 weeks earlier, based on the range of estimated mutation rates of coronavirus genomes (see Materials and Methods). This timeline puts the presence of proCoV2 in late October 2019, which is consistent with the report of a fragment of spike protein identical to Wuhan-1 in early December in Italy, among other evidence. The sequenced segment of the spike protein is short (409 bases). It does not span positions in which 49 major early variants were observed, which means that the Italian spike protein fragment can only confirm the existence of proCoV2 before the first coronavirus detection in China.
Doesn’t this just exacerbate the difficulty of the question of why there weren’t earlier large-scale outbreaks? There’s no suggestion here that what they’re calling “proCov2” was not, just like Wuhan-1, already totally competent at infecting humans.
I sort of see what you’re getting at in that if the virus was already circulating in other locations, it makes the location of the first detected outbreak in Wuhan more likely to be random; it could equally have come to prominence somewhere else. But I don’t think that’s true, because these variants were all detected in samples from Wuhan, which means that is likely where the mutations took place. Otherwise you’re positing multiple events of the importation of different strains into Wuhan.
I don’t have a great cite ready to hand here, but my recollection is that COVID is unusually reliant on superspreading events, and as a result the median number of new infections from a case is quite a bit less than the mean. That could make it more plausible that a small number of infections (from whatever source) were slowly grinding along for a while until the Huanan cluster really touched things off.
The median number of new infections is in fact zero, or at least that was the case for the original variants, who knows about these new Delta ones and whatnot.
Most of the discussion around the spread of SARS-CoV-2 has concentrated on the average number of new infections caused by each patient. Without social distancing, this reproduction number - R - is about three. But in real life, some people infect many others and others don’t spread the disease at all. In fact, the latter is the norm, Lloyd-Smith says: “The consistent pattern is that the most common number is zero. Most people do not transmit.”
But in a recent preprint, Adam Kucharski of LSHTM estimated that k [a dispersion factor; measures how much the virus clusters] for COVID-19 is as low as 0.1. “Probably about 10% of cases lead to 80% of the spread,” Kucharski says.
That could explain some puzzling aspects of this pandemic, including why the virus did not take off around the world sooner after it emerged in China, and why some very early cases elsewhere—such as one in France in late December 2019, reported on 3 May—apparently failed to ignite a wider outbreak. If k is really 0.1, then most chains of infection die out by themselves and SARS-CoV-2 needs to be introduced undetected into a new country at least four times to have an even chance of establishing itself, Kucharski says. If the Chinese epidemic was a big fire that sent sparks flying around the world, most of the sparks simply fizzled out.
1 Like
This is true, but the problem with this as an explanation for how this could have been spreading at low levels in September is that when there are no precautions in place, a “super spreader event” is one guy spending even 5 minutes inside with other people.
Really in-depth Zeynep on COVID origins in NYT. Nothing new, but a really good synthesis. Some what-now thoughts at the end too:
But even if we are denied answers, we can still learn lessons.
Perhaps the biggest one is that we were due for a bat coronavirus outbreak, one way or another, and the research showing bat coronaviruses’ ability to jump to humans was a warning not heeded.
Scientists and government officials need to weigh the benefits and dangers of how we work with bats and viruses, in the field and the lab, especially since other public health investments may do much more to prevent a pandemic. It might be more effective to institute rigorous surveillance where threatening pathogens are known to thrive, and better prepare our institutions to react quickly and transparently to the first sign of an outbreak. Research can be weighted toward response rather than prediction; these overlap but aren’t identical. Finding a dangerous virus in a cave or a petri dish might be useful, but it’s a bit like poking a bear we are trying to avoid.
Field research on bats should have been done more carefully. Bat viruses should not be studied in BSL-2 labs, and research in BSL-3 labs should be done only under the strictest caution. Bats should be treated as a serious threat in labs. Human interactions with bats should occur under strict regulation and surveillance.
…
Some scientists have proposed imposing stricter controls and a stronger risk-benefit analysis for research on pathogens that could inadvertently spark pandemics. Some research may still be worth it, and there have been proposals to move such labs outside densely populated cities.
Cooperation with China on these issues is vital, including on lab safety and outbreak surveillance. Some argue that criticizing China’s response to the pandemic and the scientific practices that might have led to it will imperil that cooperation. It’s hard to see how angry op-eds could make Chinese officials more intransigent than they already are.
People are understandably wary that these claims might demonize scientists from other countries, especially given the anti-Asian racism that has abounded. But why would perpetuating this state of events be to their benefit?
After a lab accident with anthrax bacteria in the Soviet Union in 1979 that killed dozens, leading Western scientists accepted the Soviet government’s excuses, which all turned out to be lies. That doesn’t help lead to better safety standards, including those that would benefit scientists in authoritarian countries.
…
To do this, government officials and scientists need to look at the big picture: Seek comity and truth instead of just avoiding embarrassment. Develop a framework that goes beyond blaming China, since the issues raised are truly global. And realize that the next big thing can simply mean taking great care with a lot of small details.
1 Like
Yeah, this neither fizzled out nor exploded, but circulated at a low level for about two months before breaking out in December.
The only compelling evidence for the lab leak theory was the curious proximity of the WIV to the wet market superspreader event. If we now say that ground zero was actually two months prior at an unknown location, the battle against Occam’s Razor has become even more challenging.
I think this is overstated. COVID isn’t Captain Trips. If 5 minutes of indoor contact was all you needed to reliably infect someone else, R_0 would be a lot higher than 2-2.5 (and this is for D614). The key point is that under “perfect storm” conditions, COVID can infect people terrifyingly efficiently, but a lot of things need to align right to make that happen. (Also worth noting that a lot of East Asia has developed cultural practices around mask-wearing that likely make it harder for COVID to spread–not sure about Wuhan specifically, but speculating that it’s similar there.) If there’s only a few cases, there’s high variance in how many new infections there are from generation to generation.
2 Likes
Yeah, all this. And then the first identified case in that database Bloom hacked was December 26 or something. In Italy. And it didn’t explode there until the middle of February. So you’re basically talking about the same sort of timeline in Wuhan as happened in Italy.
Although it’s not clear to me what the heck that early Italy sequenced sample actually is, if it was actually collected at the date that the database indicated, or whatever. And you for sure had documented infected Chinese tourists in mid to late January going to northern Italy and then shit didn’t really hit the fan until late February. And this was with the big boy covid strains that overwhelmed Wuhan.
The NYT has an article on the origins of COVID that is pretty good. There’s no killer new information or anything but it lays everything out well and is worth a read.
Probably the most interesting part is about a number of scientists who signed the Lancet letter denouncing lab accident theories who are now recanting:
Several scientists who signed The Lancet letter denouncing the consideration of anything but natural origins have since said they are more open to lab involvement. One, Bernard Roizman, an emeritus virologist at the University of Chicago with four honorary professorships from Chinese universities, said he was leaning toward believing there was a lab accident.
“I’m convinced that what happened is that the virus was brought to a lab, they started to work with it,” he told The Wall Street Journal, “and some sloppy individual brought it out.” He added, “They can’t admit they did something so stupid.”
Charles Calisher of Colorado State University, another signatory, recently told ABC News that “there is too much coincidence” to ignore the lab-leak theory and he now believes “it is more likely that it came out of that lab.”
Peter Palese, the virologist who wrote about the 1977 flu pandemic, said that “a lot of disturbing information has surfaced since The Lancet letter I signed” and that he wants an investigation to come up with answers.
This one is pretty good. More scientifically literate focus on what these labs were doing, also a lot of hilarious asides about the Shi-Baric frenemy-ship. Baric seems deeply committed to the position that fucking around with weird pandemicogenic bat viruses is good, but not in his rivals’ labs.
More substantively, the essay seems to say that Rand Paul was basically right about NIH funding WIV, or if he wasn’t, it’s down to something you’d call semantics (or some similar word).
1 Like
I dunno, it seems like his distinction between BSL-2 and BSL-3 is pretty reasonable.
BSL-2 is for moderately hazardous pathogens that are already endemic in the area, and relatively mild interventions are indicated: close the door, wear eye protection, dispose of waste materials in an autoclave. BSL-3 is where things get serious. It’s for pathogens that can cause serious disease through respiratory transmission, such as influenza and SARS, and the associated protocols include multiple barriers to escape. Labs are walled off by two sets of self-closing, locking doors; air is filtered; personnel use full PPE and N95 masks and are under medical surveillance.
That’s a gigantic step up in safety.
1 Like