When our immune system encounters a new virus—say, SARS-CoV-2—for the first time, some of our plasmablast B cells will release a first flush of antibodies that can stick to and even neutralize the virus. But those first antibodies are short-lived, typically don’t stick very strongly, and the cells that produce them don’t last longer than a few weeks.
As reinforcement, the immune system has a program in place to create long-lived plasma cells that secrete even better, more–tightly binding antibodies. This takes place in the germinal center, a transient, complex structure in the lymph nodes and spleen. There, different B cells gather and undergo a process that’s best described as evolution on the fast track. They rapidly proliferate while deliberately introducing random mutations to certain genes encoding antibodies, increasing the chances of producing a few clones that can squelch the virus.
A type of T cell picks and chooses progeny with promising antibodies, while killing the less useful ones or directing them back for improvement—a cycle of refinement and selection that culminates in the creation of long-lived, plasma B cells that can protect us against re-infection for years, and also memory B cells, which last for even longer. Without germinal centers, the body cannot make lasting antibody immunity to pathogens.
Some severely affected COVID-19 patients entirely lack those structures, according to a post-mortem analysis of lymph and spleen tissue published last week (August 19) in Cell. The authors posit this could be due to high levels of a certain cytokine, TNF-α, which impairs the differentiation of the follicular helper T cells that are required for these germinal center reactions. Although it’s still unclear if asymptomatic or mildly affected COVID-19 patients induce better germinal center responses, the findings help explain why antibody-based immunity to SARS-CoV-2 may be relatively short-lived.
The study “suggests an explanation for a very big problem,” says Ziv Shulman of the Weizmann Institute of Science in Israel who wasn’t involved in the study. “The strength is that it’s not a model—we’re looking at the real disease.”
As a group that has previously studied germinal centers and B cell development, it was only natural for immunologist Shiv Pillai of the Ragon Institute of Massachusetts General Hospital, MIT, and Harvard and his colleagues to investigate the sites in COVID-19 patients where germinal centers would be expected to form. They examined the thoracic lymph nodes and spleens of 11 COVID-19 patients who had died of severe infection between March and May this year at Brigham and Women’s Hospital in Boston.
All of them had these weird cells.—Shiv Pillai, Ragon Institute
Yet when they stained the tissue with different markers typical of germinal centers, they found no signs of the antibody-producing depots. This was in contrast to thoracic lymph nodes and spleens from people who had died of other causes. All of the control patients had germinal centers, likely because of ongoing immunity to harmless pathogens they carried.
“The thing you’d expect to see is loaded germinal centers,” Pillai says of the COVID-19 patients. But the absence of the structures wasn’t a complete surprise. Since April, several studies have reported antibodies in COVID-19 patients that don’t show the characteristically heavy mutational signature one would expect if they had stemmed from a germinal center. In addition, some autopsy reports of patients who died during the original SARS outbreak in 2003 found that germinal centers were absent there as well, notes Pillai, who’s on the scientific advisory board of the Massachusetts-based antibody therapeutics company Abpro and soon-to-launch Delaware-based B cell engineering startup Pulsar Biopharma.
In their tissue analyses of COVID-19 patients, Pillai’s team also documented a striking reduction of a particular kind of B cell that typically populates the germinal centers and produces the marker Bcl-6. This was accompanied by a lack of Bcl-6–bearing follicular helper T cells, which are important for kick-starting the formation of the germinal center as well as selecting promising B cells there. The team surmised that the cytokine TNF-α, which was in unusually high concentrations around the would-be location of the germinal center, could be blocking the differentiation of T cells into Bcl-6 follicular helper T cells, thereby preventing the formation of germinal centers. In mouse models of malaria in which germinal center reactions are defective, it’s known that blocking the cytokine can restore functional germinal centers.
The immune systems of these patients may be taking an alternate route to create antibodies. The team found a preponderance of another type of B cell in the lymph nodes and spleens of COVID-19 patients, so-called double-negative B cells, which have been linked to autoimmunity and other pathological conditions. These are capable of creating antibody-secreting cells, but not the kind needed for long-term antibody immunity, notes Ankur Singh, an immune engineer at Georgia Tech who wasn’t involved in the research. “It’s a very useful study in terms of understanding how the virus is impacting critically ill people,” he adds.
To Shulman, the absence of germinal centers in COVID-19 patients who died—whose immune systems weren’t capable of mounting an effective immune response—isn’t surprising. A key question is, “can we learn something from that about patients who don’t die or even asymptomatic [people] . . . about long-lasting immunity?”
Pillai’s team also conducted an analysis of blood from 68 living COVID-19 patients, spanning moderately ill, severely ill, and recovered patients. Upon analyzing the patients’ cells, they found that the double-negative cells and other B cell subtypes that are typically generated outside the germinal center—including ones that were specific to SARS-CoV-2—were prominent across moderately and severely affected patients. “All of them had these weird cells,” Pillai says.
Although not proof that the inability to form germinal centers is a widespread phenomenon in COVID-19, Pillai says he suspects that could be possible, even in mild cases. Most SARS-CoV-2 antibodies documented so far exhibit a very small degree of hypermutation, also among mild cases, which is suggestive that they didn’t arise from a germinal center. If a failure to produce germinal centers is common, then so too would be re-infection with SARS-CoV-2, and that could have implications for the ability to reach herd immunity through natural infection, the authors suggest. However, it wouldn’t have any bearing on a vaccine, because vaccination typically generates good germinal centers, Pillai says.
COVID-19 is a continuum of severity and symptoms, Singh says, and one would expect something like germinal center formation to fall somewhere on that spectrum as well. “It’s quite possible that patients that have less severe disease may induce better germinal center reactions.” In addition, some research has found a few mutated antibodies among the predominantly unmutated ones in people who have recovered from COVID-19, which means that “germinal center activity must take place to some extent,” Shulman adds.
Both Singh and Shulman say they would like to see more research on the underlying mechanisms. It’s not clear, for instance, why TNF-α—one of several cytokines elevated in COVID-19—might have this effect, and whether it’s the only cause of absent germinal centers. “I think that conclusion cannot be made from this particular study,” Singh says. It’s also very hard to disentangle whether it’s a defect in the follicular T helper cells, or a B cell defect, that impairs the germinal center response, Shulman adds. “It’s very hard to say because the cells depend on one another.”
Other mechanisms are possible, Shulman notes. The antibodies initially formed in COVID-19 patients, although they don’t stem from the germinal center, are relatively high-affinity antibodies and are often capable of neutralizing the virus. It’s feasible that because the initial antibodies are good enough, the immune system doesn’t proceed with germinal center reactions, and makes do with antibodies that don’t provide long-lasting protection. Figuring out how the germinal center is disrupted is something that requires animal studies, Shulman notes.
To that end, Pillai is currently trying to model mild COVID-19 in mice to investigate these questions. As for the question of long-term immunity, he says, “time will tell. We have to follow patients up for a while, and that’s what we’ll do.”
N. Kaneko et al., “Loss of Bcl-6-expressing T follicular helper cells and germinal centers in COVID-19,” Cell, doi:10.1016/j.cell.2020.08/025