Thread by @Ayjchan

Many recent articles, interviewing EcoHealth experts, are advocating for more funding to search for the source of SARS-CoV-2 and novel pathogens. In this tweetorial, I'm diving into their bat cave CoV sampling pipeline to understand how this helps us to prepare against pandemics.

One of the best scientific papers for understanding this process is the Nov 2017 PLoS Pathogens report of EcoHealth & WIV’s 5-year surveillance of SARS-related CoVs (SARSr-CoVs) in a bat cave, 60km from Kunming city, in Yunnan province.
journals.plos.org/plospathogens/…

Co-authored by major players in the field, including P Daszak, LF Wang, J Cui, and ZL Shi, the study found that SARSr-CoVs, e.g., WIV1 and WIV16, in Yunnan were more closely related to the 2003 SARS (emerged in humans in Guangdong province) than CoVs from other regions in China.

Although the team could not definitively answer WHERE the 2003 SARS came from, they inferred that the progenitor originated from recombination within horseshoe bats (hosts) in the Spike and ORF8 (2 genes in the 30kb CoV genome).

Lack of data on animal trade and lack of uniform sampling of bats in China made it difficult to determine whether other caves closer to the site of outbreak could have been the source of the 2003 SARS virus. Or whether SARS had been ferried from Yunnan to Guangdong (1400km).

However, they effectively concluded that there is risk of spillover and another SARS outbreak. Within months, they published a letter to editor describing how 2.7% of the sampled persons living near Yunnan bat caves had antibodies to SARSr-CoV proteins (0% in Wuhan).

Of the 6 individuals who tested positive, all had bats flying in their villages. Only 1 had traveled out of Yunnan in the past year to Shenzhen (Guangdong).
ncbi.nlm.nih.gov/pmc/articles/P…

It is unclear what impact these scientific findings had at the time. Were policies enacted to prevent spillover of viruses from bats into humans in Yunnan province? and trafficking of human-transmissible CoVs out of Yunnan?

They cited a 2016 study by R Baric which evaluated spillover potential by creating and testing chimeric CoVs. They found that WIV1 CoV from bats can directly infect humans but require “additional adaptation for epidemic disease”.
pubmed.ncbi.nlm.nih.gov/26976607/

Baric’s team tested WIV1 as well as a mouse-adapted WIV1 carrying a SARS spike. They generated these (chimeric) genomes in their 2003 and 2016 papers using the same approach: Class IIS enzymes that stitch genomes together leaving no scar or trace.
ncbi.nlm.nih.gov/pmc/articles/P…

Interestingly, the S1/S2 FCS out-of-frame PRRA insertion in SARS-CoV-2 generates such a Class IIS restriction enzyme site (BsaXI). This BsaXI site only exists because the PRRA insertion is out-of-frame, and requires a Proline (P) upstream of the RRA.

In other words, if the insertion was in frame OR if a different residue was upstream of RRA, the BsaXI site would not exist.

That is not to say that this is evidence for a natural or unnatural origin of the SARS-CoV-2 S1/S2 FCS. It is only an observation that a BsaXI Class IIS site overlaps exactly at the S1/S2 FCS insertion site. Several CoVs have ~8-10 BsaXI sites in their genomes.

Baric's group describe this cloning approach: "primer pairs were designed that contained a class IIS restriction enzyme, like AarI... the AarI junctions were designed to seamlessly link consensus fragments, resulting in the production of a full-length cDNA fragment"

So, back to the PLoS Pathogens paper by EcoHealth and WIV, what do you actually do once you sample bats for novel CoVs?

(1) Test bat samples (swabs or feces) for the presence of CoVs by amplifying a 440-nt RdRp region in the genome that is conserved among all known alpha and beta CoVs. Sequence the amplicons to confirm CoV presence (check against false positives).

(2) Identify the host species - what type of bat is this CoV from? Rhinolophus sinicus (Rs), ferrumequinum (Rf), affinis (Ra) etc.

(3) Characterize genetic diversity of the CoVs at the most variable region of the Spike gene: the receptor-binding domain (RBD). This region has a major impact on what host species the CoV can infect.

In the 2017 PLoS Pathogens paper, the RBD sequences could be divided into 2 clades. And in two of their bat samples, CoVs from the 2 clades could be detected (co-infection scenario).

(5) Novel strains are selected for full-length genomic sequencing based on sample abundance, RBD genotype, and sampling time.

This was accomplished by amplifying, from isolated virus RNA, overlapping genomic fragments with degenerate SARSr-CoV primers and primers derived from previous round of sequencing.

Within a single cave, the team managed to obtain 15 SARSr-CoV genomes all sharing 92-99.9% nucleotide sequence identity.

The 2017 paper paid attention to the constitution of the Spike gene: SARSr-CoVs share high identity in the S2 region, but not the upstream S1 region. S1 was used to divide the CoVs into more subgroups.

The group also noticed that strains carrying the Spike NTD similar to that of SARS were only found in 2013, although similar RBD were detected in different seasons throughout the 5 years.

(4) Characterize genetic diversity of the CoVs are another highly variable gene: ORF8, which suffered a 29-nt deletion in the 2003 SARS outbreak. *Orf3 is the next area of interest and has been found to co-evolve with the Spike.

(6) Recombination and phylogenetic analysis: determine the regions/genes in which the genomes could have recombined within the community of viruses sampled at the cave and closely related viruses (2003 SARS).

(7) Virus isolation: apply virus sample to Vero E6 (Green Monkey Kidney) cells to see if more virus particles can be cultured.

(7b) Alternatively, take just the Spike gene from the novel CoV and insert it into a closely related CoV backbone (WIV1 in this case) to try to generate virus particles in Vero E6 cells.

(8) ACE2 cell entry assay: Once you have virus particles (meaning that these can replicate in Vero cells), test on human cells with or without human ACE2 receptor expression - can the virus enter the human cells via ACE2 and replicate?

(4b) Check Orf8 protein activation of ATF6: as simple as transfecting human cells with just the ORF8 gene from different CoVs.

(4c) Check Orf8 induction of apoptosis (cell death): ~10% of human cells expressing ORF8a from SARS or SARSr-CoV died.

With this pipeline in context, what did the WIV & EcoHealth do with the closest related CoV to SARS-CoV-2? Which was discovered in July 2013 in a cave in Mojiang county, also in Yunnan province.

RaTG13 was isolated from a bat in 2013, but was only described in 2020 by Shi's group after the SARS-CoV-2 outbreak: nature.com/articles/s4158…

"... a short region of RNA-dependent RNA polymerase (RdRp) from a bat coronavirus (BatCoV RaTG13)—which was previously detected in Rhinolophus affinis from Yunnan province—showed high sequence identity to 2019-nCoV. We carried out full-length sequencing on this RNA sample"

Although Shi's 2020 paper did not cite their original paper that described their discovery of RaTG13, another group from Wuhan did so for them: ncbi.nlm.nih.gov/pmc/articles/P…

Was RaTG13 characterized after being discovered in 2013? According to Vox, Daszak said "No one [in Wuhan] cultured viruses from those samples that were 20 percent different, i.e., no one had SARS-CoV-2 in culture." vox.com/2020/4/23/2122…

On May 19 of this year, someone from the WIV uploaded sequencing data of RaTG13 dating back to 2017 and 2018: trace.ncbi.nlm.nih.gov/Traces/sra/?ru…

When the sequencing reads are mapped to the RaTG13 genome, steps (1), (3), and (4) can be observed. Although the Spike (step 3) was only sequenced in late 2018.

In this regard, the protocol for characterizing novel SARSr-CoVs was being followed, even if the experiments (steps 5-8) have not been performed/published.

Realized I used the wrong wording here. According to both WIV and EcoHealth, RaTG13 was never isolated aka cultured. Although would be useful to know, especially after SARS-CoV-2 broke out in December, whether RaTG13 can replicate in Vero (primate) cells and infect human cells.