Read on Twitter
Excited to share this new story from our lab, characterizing the structure and function of the 3a ion channel of SARS-CoV-2!
https://www.biorxiv.org/node/1372064
https://www.biorxiv.org/node/1372064
This story is part of a larger collaboration between our lab, and those of Diana Bautista and Hillel Adesnik focused on the SARS-CoV-2 ion channels. We were fortunate to receive funding from FastGrants early in the process. ( https://fastgrants.org/ )
It's worth mentioning that this project was only started ~ 3 months ago, and the rapid pace is largely thanks to the substantial effort of @kerndm who did all of the molecular biology, and biochemistry (along with much more) throughout this pandemic.
My role was 100% remote as I sheltered-in-place, and was focused on the cryo-EM processing of the apo dataset. While in parallel @kerndm processed the emodin added and tetramer datasets. Really fun working together to move as quickly as possible!
Tetramer and emodin you ask? Beyond the apo structure we also resolved the structure of tetrameric 3a (albeit at lower resolution, ~6.5 Å).
Emodin was previously suggested to inhibit SARS-COV-1 3a ( https://doi.org/10.1016/j.antiviral.2011.02.008) however when David added 100uM to grids, we see no resulting density or change in the protein conformation. This matches twitterless-Ben's recordings, finding no inhib. from emodin in our hands.
All three EM datasets have been uploaded to EMPIAR!
One thing that immediately jumps out from the structure is the location of Q57, shown here from a view "above" the membrane.
We found it especially intriguing that a Q57H variant is observed in ~25% of the sequenced SARS-CoV-2 genomes.
However, when David purified the mutant protein, reconstituted proteoliposomes, and Ben patched them -- we saw no difference from wild-type protein.
Relatedly, see this interesting paper from Berkeley colleague @AprilWei001 and team: https://www.biorxiv.org/content/10.1101/2020.05.21.109280v2
When we map other observed variants on the structure, many seem unlikely to impact channel function, however 3 (colored pink here) sit at the proposed tetramer interface.
We think that part of the N-terminus could account for this density (colored orange) that sits right above the channel. David made, purified and reconstituted a ΔN 3a -- and Ben's careful recordings again showed no difference from wild-type channel.
However, ΔN 3a had clear increased plasma membrane localization. (These images are thanks to Savitha Sridharan from Hillel's lab!)
I also want to give a shout-out to Jonathan and Dan from UC Berkeley's cryo-EM facility. Due to pandemic safety measures, we were unable to collect the data ourselves and could not have done this without their help.
Finally, Steve really dug into the hunt for structurally related 3a proteins. Interestingly, he found that all coronaviruses with 3a structural homologs appear to be descended from the bat gene pool! 
We would love to hear what you think, please reach out!
