Our preprint is up on bioRxiv! We discuss key findings about the adaptation of SARS-CoV-2 for humans, scenarios for cross species transmission (pangolins, Wuhan Huanan market), and measures to prevent re-emergence of COVID-19.
First Tweetorial. We first noticed that, over similar 3 month periods, SARS-CoV-2 (COVID-19) is less genetically diverse than SARS-CoV (2003). From Jan-Mar 2003, SARS-CoV genomes diverged by up to 85 nt. In comparison, from Dec-Mar 2020, SARS-CoV-2 genomes diverge by 15-25 nt. 
SARS-CoV-2 (red) resembles late epidemic SARS-CoV (yellow) after adaptive mutations had developed in the early epidemic phase of SARS-CoV (blue). Attn to Orf1a, Orf3a, and the spike, which were under positive selection to adapt to new hosts in the 2002-2004 SARS outbreaks.
The exceedingly high level of shared identity among SARS-CoV-2 isolates makes it impossible to accurately model selective pressure on each gene. This will become more feasible after many more mutations occur and, ideally, when SARS-CoV-2 from intermediate host species are found.
Focusing on the SARS-CoV spike, which binds to the host receptor (ACE2) to allow cell entry, you can see the numerous adaptive mutations that evolved and eventually dominated the late phase of the epidemic. The earliest isolate is at the bottom row.
In contrast, most of the non-synonymous mutations in the SARS-CoV-2 spike are scattered across the gene and have not been reported to confer adaptive benefit. Despite infecting at least 3.3 million people to date, there is no evidence of a more virulent strain emerging.
A comparison of the spikes from SARS-CoV (2003) and SARS-CoV-2 (2020) reveal that the SARS2 spike binds 10-20x more strongly to human ACE2 receptor and bat ACE2 receptor. SARS2 spike also has superior plasma membrane fusion capability.
The single site of notable entropy in the SARS2 spike is D614G, which sits outside of the receptor binding domain and is not predicted to significantly impact spike structure or function (it perturbs one hydrogen bond). However, D614G is in the middle of a B-cell epitope.
Ultimately, these observations suggest that by the time SARS-CoV-2 (COVID-19) was first detected in late 2019, it was already pre-adapted for human transmission - to an extent more similar to late epidemic SARS-CoV rather than early-to-mid epidemic SARS-CoV.
Reduced genetic diversity can stem from a bottleneck event such as in the case of late epidemic SARS-CoV where a single superspreader at the Metropole Hotel in Hong Kong progenated the vast majority of the international SARS cases in 2003.
However, if SARS-CoV-2 (COVID-19) isolates today all stem from a bottleneck, where are the sibling branches evolving from a less human-adapted version of the virus? This suggests a single introduction of a human-adapted SARS-CoV-2 into humans in late 2019.
There were two back-to-back outbreaks of SARS in 2002-2004 arising from separate civet-to-human transmission events. The first outbreak emerged in late 2002 and ended in August, 2003. The second outbreak emerged in late 2003 and was rapidly suppressed.
To learn from the past and prevent near future re-emergence of SARS-CoV-2 from an unidentified source, it is vital to consider all routes by which SARS2 could have adapted for human transmission. It would be dangerous if there are lingering pools of human-adapted SARS2.
No definitive evidence supports or rules out any of these origins: (1) SARS2 crossed into humans, circulated undetected while adapting to humans (2) SARS2 was human-adapted in another species b4 crossing to humans (3) SARS2 adapted to humans in a lab without genetic engineering.
Based on the first two scenarios, what is known about possible intermediate hosts and SARS-CoV-2 species tropism?
Speculations that pangolins are the likely intermediate host stemmed from the discovery of a pangolin CoV that shares 95.4% S amino acid identity and six key RBD residues with SARS-CoV-2. Since then, another closely related lineage of pangolin CoVs has been identified.
However, the unique polybasic furin cleavage site in the SARS2 S is not found in pangolin CoVs. SARS2 is not a recent recombinant of known CoVs. The CoV most closely related is RaTG13, a bat CoV identified at the Wuhan Institute of Virology, isolated from Yunnan, China.
SARS2 shares 96%, 84%, and 80% genome identity with the bat RaTG13, pangolin MP789, and 2003 SARS-CoV, respectively. No evidence points to the human-adaptation of SARS2 in pangolins or transmission of SARS2 from pangolins to humans.
It is plausible for SARS-CoV-2 spike to have evolved its broad species tropism (ability to bind to the ACE2 receptor across diverse species) naturally in bats or a wide range of intermediate species.
SARS2 spike could bind to ACE2 from more than 100 species. The spike of RaTG13, similar to SARS2, binds to both bat and human ACE2. The spike of MERS-CoV binds to receptors from humans, camels, and bats - adapting to semi-permissive host receptors in 3 passages in cell culture.
Broad Host Range of SARS-CoV-2 Predicted by Comparative and Structural Analysis of ACE2 in Vertebrates
biorxiv.org/content/10.110…
biorxiv.org/content/10.110…
These findings collectively suggest that some coronaviruses in nature are evolving spikes that can bind at an optimal level to the same receptor across diverse species, potentially by interfacing with highly conserved parts of the receptor.
CoV sampling from more species - to avoid bias stemming from the focused scrutiny of Malayan pangolins - will provide us with a better grasp of the range of animal species that harbor CoVs with similar RBDs to SARS-CoV-2, as well as the natural diversity of bat CoVs.
Did SARS-CoV-2 originate from the Wuhan Huanan seafood market? Scientists and the public continue to debate. According to the Chinese CDC website (Apr 27, 2020), SARS2 was detected in environmental samples from the market and thus likely originated from animals sold there.
In contrast to the thorough and swift animal sampling in response to the 2002-2004 SARS-CoV outbreaks to identify intermediate hosts, no animal sampling prior to the shut down and sanitization of Huanan market was reported. Details about the environmental sampling are sparse.
Only 4 out of the 515 environmental samples have passable coverage of SARS2 genomes for analysis. Compared to a human SARS2 isolate, the market samples all shared >99.9% genome and spike gene identity.
In 2002-2004, only SARS isolates collected in a narrow window of time within the same species shared >99.9% identity. This makes it unlikely for the Huanan market isolates to have come from an intermediate animal host; likely from SARS2-infected humans who visited the market.
Furthermore, if SARS2 originated from an animal at Huanan market, it is curious that no other SARS2-like viruses have been detected at the market. Presumably the intermediate host would've carried a pool of SARS2-like progenitors that could jump into humans and other animals.
If intermediate animal hosts of SARS-CoV-2 were present at the market, no evidence remains in the genetic samples available.
Without evidence to rule out SARS-CoV-2 adaptation in an intermediate host, humans, or a lab, we need to take precautions against each scenario to prevent re-emergence. Particularly if a pool of human-adapted SARS-CoV-2 progenitors still exist in animals.
The response to the first SARS-CoV outbreak deployed the following strategies that were key to detecting SARS-CoV adaptation to humans and cross-species transmission, and could be re-applied in today’s outbreak to swiftly eliminate progenitor pools.
(i) Sampling animals from markets, farms, and wild populations for SARS-CoV-2-like viruses.
(ii) Checking human samples banked in early/mid 2019 for SARS2-like viruses or SARS2-reactive antibodies to detect precursors circulating in humans. SARS2 isolates from Wuhan, particularly early isolates, could identify branches originating from a less human-adapted progenitor.
(iii) Evaluating the over- or underrepresentation of food handlers and animal traders among the index cases to determine if SARS-CoV-2 precursors may have been circulating in the animal trading community.
Meanwhile, it would be safer to more extensively limit human activity that leads to frequent or prolonged contact with wild animals and their habitats.
@RosarioP70 I recommend this paper- they analyzed hundreds of animal samples for the ACE2 gene and predicted computationally which ones can bind the SARS2 spike. For now they have not tested animals or proteins in the lab, so we cannot tell yet which animals can host SARS2.
@emilykwong1234 @gbrumfiel Heard your first CoV podcast as a bonus on Up First today + found your newest podcast about zoonotic transmission. My colleagues & I would be interested in chatting with you about our findings about SARS2 adaptation if you're making another episode!
