A lot of people have had one or more of the coronaviruses that are listed on the CDC page (229E, NL63, OC43, HKU1). But none of these are in the exact same family as the current beast – the first two are alpha-coronaviruses and are fairly closely related to each other. The second two are in another genus, beta-coronaviruses, and are also pretty closely related to each other, but they’re off in a different lineage inside the beta-coronaviridae compared to the SARS-type coronaviruses like the current one. All of these things have “spike” proteins decorating them, but the spikes themselves vary in sequence, enough so that some of them have found completely different surface proteins to use for viral entry, as opposed to the SARS ones going for the ACE2 protein.
Still, immunology being what it is, the question has been open whether the B-cell and T-cell memory of past infections with these other coronviruses might give a person some protection against the current one. I’ve wondered about that here on the blog myself. It’s not at all a crazy idea, because what we have seen is that there are people out there who with cross-reactive antibodies that can bind to the pandemic coronavirus, some of these in blood samples from well before the current one started going through the human population. but (until now) we’ve lacked enough hard data to say.
Here’s the MedrXiv version of the paper under discussion, and here’s the version coming out now in Cell. The authors looked at 431 pre-pandemic blood samples, and compared them to 251 samples from people who have been infected in the current outbreak and recovered, as well as analyzing antibody profiles in people who are currently hospitalized. What they’ve found is first, that most people have indeed been infected with one or more of the “garden-variety” coronaviruses. The pre-pandemic samples show plenty of antibody responses to these. Second, about 20% of these patients raised antibodies that do cross-reaction with the Spike or nucleocapsid proteins of the current pandemic coronavirus. And what’s more, levels of such antibodies are elevated when a person in this group gets infected with SARS-Cov2: the immune system memory (as present in these patients’ B cells) responds by increasing production of the antibodies to the previous coronaviruses.
But here’s the key part: “cross-react” does not mean “neutralize” and it does not mean “provide protection from”. These antibodies may or may not have been neutralizing against the other coronaviruses, but they don’t seem to have any such effect on the current one. And in keeping with that, having such cross-reactive antibodies seems to provide no protection against catching SARS-Cov2 or against being hospitalized with it if you do. There’s no difference in the infection/hospitalization rates of the people who had cross-reactive coronavirus serum antibodies ready to go versus those who didn’t. They’re basically useless.
Now, you can still make an argument that the T cell component of immunity might provide some protection after a previous coronavirus infection. The current study didn’t address this directly, but after these results, it’s at least less likely that that’s happening. The authors make a note of this, and also note that pre-existing mucosal antibodies might exert a protective effect (which this study didn’t examine, either). But prior circulating human coronavirus antibodies, even ones that can bind to the current one – those it looks like we can rule out. Which is too bad.
