Before we begin, here are two bigcaveats: First, both of these articles are in preprint. They may still be subject to change, or even withdrawal. Second, each of these papers contains information that is beyond my ability to properly evaluate and verify. To a large extent, Im accepting the assertionsof the authors, especially when it comes to the Cell paper, which requires a knowledge of immune systembehavior that definitely does not linger from classes I last took in 1978. (Probably because it wasnt even understood in 1978.) Okay, lets continue:
The initial paperfrom Cellappeared on May 7 and comes from a team from UNC, UCSD, the La Jolla Institute for Immunology, and Mt. Sinai. They looked into an aspect of the immune system that isas important as antibodies, butgets far less coverage in the press: T cells.T cells are a type of white blood cell, or lymphocyte, that come from the thymus. Unlike some white blood cells, T cells are adaptive. That is, past exposures toproteins leave behind T cells that are prepped to respond if that protein is encountereda second time.
The team behind the Cell paper looked at two different kinds of T cells:CD8 cells, which go after cells that have been infected by a virus, and CD4 helper cells, whose main task is kicking off other parts of the immune responseincluding signaling for production of more CD8 cells. These cells arent just behind a big part of the bodys effort to suppressviral infections, theyre also at the heart of acytokine storm response that can cause a lot ofdamageor even deathwhen the immune system overreacts to an intruder.
With that in mind, the team looked first at patients who had tested positive for active cases of COVID-19. In those patients, they found CD4 T cells with markers for the protein found on the spike of the SARS-CoV-2 virus, which isa critical component of how it enters cells. They also found markers for a number of other proteins including ones labelled M and N (Im not going to bother to try to explainwhat those do). All three were found in 100% of the exposed samples.When it came to the CD8 T cells, the spike and M protein were again strongly indicated in all samples. N was a little less so.
This research alone has strong implications for anyone developing a COVID-19 vaccinebecause it suggests that these three proteins alone are likely enough for developing a strong, distinctive immune response. To be sure of a strong response, it would be good to use all three, and possiblya few others. The universality of these T cells is another good sign that long-term immunity to COVID-19 is a likely result of infection, and that effective vaccines are possible. So all good news.
But that was only the first step. The team then looked at samples of individuals who had definitely not been exposed to COVID-19, mostly because the blood samples were taking in 2018 or sooner when there was no COVID-19 virus (at least not in the human population). And here comes the moment that will (or already has) generate a thousand confusing tweets:40% to 60% of unexposed samples had CD4 Tcells that responded to the novelcoronavirus. All of these samples showed indicators for a pair of common cold viruses that cause upper respiratory infections:HCoV-OC43 and HCoV-NL63. (The HCoV in both of these just stands for human coronavirus.)Both of these viruses are globally endemic, meaning they are widespread and common.
In particular, non-spike specific responses were above the limit of detection in 50% of the unexposed samples. Thats because while these other coronaviruses dont share the same protein that SARS-CoV-2 uses to penetrate cells with its spike, they do have other proteins, like M and N, in common.Most discussion in the mediaand much of the talk about vaccineshas focused on the spike protein, but the reaction to the M protein was just as strong in the samples examined by this team, with reactions to N not far behind. So it may be that these proteins are just as effectiveand importantas the spike protein when it comes to an effective immune response.
(If youre still with me at this point, grab a coffee or at least a deep breath, because were halfway. But trust me, its interesting.)
The second paper arrived on the preprint sitebioRxivon May 15. And while the Cell team might have been circumspect in discussing the possible implications of their results, the bioRxiv group goes for the jugular right in the title of this one: Pre-existing andde novohumoral immunity to SARS-CoV-2 in humans. Or, in slightly more quotidian language: They looked at the antibodies found in people who had not been exposed to the novelcoronavirus and compared themto antibodies from those who had been exposed.
The list of authors here includesno fewerthan 32 names, and the number of institutions involved is almost as varied, with a core team from University College, London. This sprawling team examinedthe contention that endemic humancoronaviruses, like the ones from the Cell paper, might provide Immune cross-reactivity for SARS-CoV-2. Instead of looking at T cells, they went directly after antibodies and they found them.
The work of the antibody group was based around a theory that infection by a number of human coronaviruses can provide a level of cross-protection, albeit transient against other human coronaviruses. That is, catching one kind of cold virus may provide temporary protection against another kind, even if they are not the same virus, because the first virus may generate some antibodies that are still reactive against the second.
On an antibody basis, the spike protein was broken down into two subunits.One of these involved how the virus attached to the cells, while the other was more involved with how the virus entered the cell after attachment. Those samples from patients exposed to SARS-CoV-2 had a strong reaction for both subunits.When it came to the samples from those who had not been exposed, there was no sign of antibodies to that first partthe attachment subunit. However,antibodies to the second subunit were detectable. So were antibodies to some of the other portions of SARS-CoV-2, including the antibody counterpart to the N protein in the first paper. This was particularly truewhen the samples came fromindividuals with recent HCoV infection.
The conclusion of the antibody team isnt only that patients who have had other human coronaviruses possibly share some immunity in the technical sense of just having antibodies, but this could havea very real effect on the outcome of cases. For example, they take note of a 60-year-old patientwho tested positive for COVID-19 but whose case remained mild and whose antibodies, even after infection, looked more like that of a patient who had never been exposed. In fact, the patient appeared to be chronically infectedhe hadsporadic positive results to COVID-19 testing for over a month while never showing more than mild symptoms. The antibody team took this as confirmation of what they had been hypothesizing: Existing antibodies to HCoVs that shared some components with SARS-CoV-2 left behind at least some level of transient immunity that protected against the development ofmore severedisease.
In particular, their results showed a strong response ofSARS-CoV-2 antibodies in people who had recently been infected by the cold virusHCoV-OC43one of the two viruses that the T cell team pointed out as generating a response that overlapped with that ofthe novel coronavirus.
Both papers suggest that patients who havehad other human coronavirusesand in particular those who have recently had a chest cold caused by human coronavirusHCoV-OC43have immune systems that are to some degree primed to fight off an infection by SARS-CoV-2. A study of that cold virusfound it was generally connected to a mild upper respiratory infection which is a lot better than having severe COVID-19.
This virus, and others that share similar proteins and structures, are endemic and common. Infection by these viruses may be a major factor in why about 85% of those infected with COVID-19 have relatively mild caseswhile around 50% of that 85% appear to have cases that are very mild or asymptomatic.
Testing of COVID-19 patients has indicated that a percentage of themsomething on the order of 15% in at least two studieshave low levels of SARS-CoV-2-specific antibodies. These results have been correlated with those who have had mild cases, and may also be connected to those who have had recent infections by other human coronaviruses and acquired a higher level of transient immunity.
Children may be more immune to COVID-19 at least in part because they are more likely to have a recent infection byHCoV-OC43, or a related coronavirus.
The shared antibodies with other human coronaviruses may be part of the reason that antibody tests, including those conducted directly on patients and those looking at sources like antibodies found in waste, seem to so often suggest a higher level of infection than might be indicated by testing or medical outcomes.
This might also explain why some group exposures form a hot spot while others dontin some cases, there may have been some herd immunity in effect, just from chance clusters of people carrying existing transient immunity.
None of this is certainin thisconclusionIve taken things at least half a logical leap beyond the position of either paper. Butif substantiated, these results could go a long way toward explaining why the immune response to COVID-19 is so extremely varied.
These papersalso strongly suggest that some people have at least a partial safety shield when it comes to developing a severe case of COVID-19. That cough you had back in December or January may not have been COVID-19, but it may save you from catching COVID-19.
But pleasedont test that.