Thread: 1) OK, so I promised to explain the manuscript that we just put on medRxiv medrxiv.org/content/10.110…
2) First, I wanted to thank our awesome collaborators @VivianaSimonLab at Sinai, @Olli_Vapalahti @hepojoki at University of Helsinki and @kedzierskalab at University of Melbourne. This would not have been possible without them.
3) Now, typically, when we get infected with a virus, we make antibody responses, especially against proteins on the surface of the virus. Often these antibodies can neutralize the virus and protect us from getting infected again.
4) The main target on the surface of most coronviruses is the spike protein, or S. below you see a model of the virus and a visualization of a crystal structure of the spike of SARS-CoV-2 (solved by @veeslerlab - heroes!). 
5) Now, in order to detect antibodies, we use an ELISA or enzyme-linked immunosorbent assay. For this, an antigen - in this case the spike protein - is coated on a sticky plastic plate (a miniaturized format). Then we let serum from patients react with it and can detect that.
6) So, first, we needed the spike antigen. We don't want to sue the virus for various reasons. So we make recombinant antigen. We can do that in insect cells or in mammalian cells in cell culture. We made two versions: a soluble version of the full spike trimer.....
7) ...and the receptor binding domain (RBD) which is part of the spike on its own (the red part in #4). Then we run them on a gel to see if they look OK and purified. i=insect, m=mammalian. Here are the recombinant RBD proteins. They turned out very nicely.
8) And here are the full length proteins. We added several stabilizing mutations. nevertheless, we got two bands. One full length version and a degradation product. We don't know why but think this is not important for assay.
9) So we had the following sera from controls (not exposed to COVID19) and from COVID19 patients.The controls are from different age groups. Also, we had serum from a person with a confirmed NL63 infection. NL63 is a human CoV that causes common cold and uses the same receptor...
10).....as SARS-CoV-2, namely ACE2. We though if we get any crossreacivity to the spike of SARS-CoV-2, then with this sample. So we ran our ELISAs. For this the serum is diluted out and you get these curves. The higher up on the y-axis, the more reactivity. We used all four......
11)....proteins as substrate to compare. Reactivity was a little lower to RBD than the full length spike. But it was clearly possible to distinguish sera from COID19 patients (red) and sera from controls (black), including the NL63 serum (green).
12) We can quantify the area under the curve to make this easier to grasp. Here we saw more background reactivity with the insect cell derived proteins than the mammalian cell derived ones.
13) I won't go into details about isotypes of mAbs we found in the positives, but that was also pretty interesting.
14) Now, what does this all mean? A) With this assay we can figure out who was infected and who wasn't. That means we can determine the true infection rate and infection fatality rate. B) We can use the assay to screen for people who seroconverted and are now immune.....
15)...and they can donate their serum and it can maybe used to treat patients. C) We can test health care workers and ask the ones who are already immune to work with infectious patients. In that way the virus is not easily spread to colleagues or other patients. And D) ....
16)...we can now use this assay to better study how our immune response reacts to the virus. And then there are two more take home messages that are important: First, it looks like we are all naive, meaning we have no immunity whatsoever to SARS-CoV-2. That would....
17)....explain why it spreads so quickly. And second, it means we make an immune response to the spike. Antibodies to the RBD domain are often neutralizing, and it is likely (but needs to be confirmed) that once the antibody response sets in, we become protected.
18) Please keep in mind that these conclusions are preliminary and based on small numbers. Larger studies to confirm this are needed and ongoing. We have started to share the reagents globally and hope that this or similar assays can be set up in many places.
19) Finally, I want to thank the student who took the lead on this, Fatima Amanat as well as my whole group of dedicated students, postdocs, techs and assistant professors who dropped all their beloved influenza work to help out with creating tools to fight SARS-CoV-2.
