ADVERTISEMENT
ADVERTISEMENT

Can We Predict How Well Someone Will Respond to a Vaccine?

Researchers find signatures pre- and post-vaccination that correlate with a more robust immune response. 

person in white jacket putting bandaid on arm of child
A black and white headshot
Natalia Mesa

Natalia Mesa was previously an intern at The Scientist and now freelances. She has a PhD in neuroscience from the University of Washington and a bachelor’s in biological sciences from Cornell University.

View full profile.


Learn about our editorial policies.

ABOVE: © iStock.com, SDI Productions

According to past research, vaccines work better for some people than others. Factors such as age, diet, and sex may influence a vaccine’s efficacy, but genetics is also thought to play a role. As of now, scientists don’t have a way to know beforehand how well a jab will stimulate any individual’s immune system.

Two papers published on October 31 in Nature Immunology bring together data from more than two dozen studies to identify the genes that determine how well vaccines in general rev up our immune systems. These findings could be applied to create better, more protective vaccines, researchers suggest.  

Previously, researchers have tried to answer the question of what genes underlie responsiveness one vaccine at a time, by taking a snapshot of the genes that individuals express before and after a particular shot. “But what’s been missing in the field is to ask: Is there a universal signature that is common to all vaccines that can be used to predict vaccine immunity?” says Stanford University School of Medicine immunologist Bali Pulendran, a coauthor on both of the papers.

The studies were spearheaded by two different labs within a cohort of researchers called the Human Immunology Project Consortium. One study, a collaboration between Pulendran’s group and immunologist Steve Kleinstein’s lab at Yale School of Medicine, found that a number of vaccines elicit a common transcriptional signature in immune cells shortly after administration, though the timescales of the signatures differ. This common signature could potentially be used to determine whether someone will generate antibodies to a vaccine.

The other study, a collaboration between immunologist Rafick-Pierre Sékaly’s lab at Emory University and Kleinstein’s lab, found that people can be categorized in one of three ways based on which inflammation-related genes their immune cells produce before vaccination. The study found that higher numbers of inflammatory markers correlate with a stronger immune response.

Compiling the data

Both studies compiled data from 25 previous studies, which included a total of 2,979 samples from 820 volunteers between 18 and 55 years old. While the studies had different endpoints, participants all received one of 13 vaccines and donated blood samples before vaccination and up to 70 days after. All of the studies measured the gene expression profile of the participants’ blood or peripheral blood mononuclear cells, a subset of immune cells in the blood, and monitored their antibody concentrations for up to a month postvaccination. 

“To our knowledge, this is the largest compilation of transcriptional data on vaccine responses,” says Pulendran. 

Mayte Suárez-Fariñas, a biostatistician at Icahn School of Medicine at Mount Sinai, who coauthored both papers as well as a previous paper that described the creation of the aggregate vaccine dataset, says that the biggest challenge was incorporating data from studies that were designed differently, with different confounders, “without damaging the data.” Some data on factors that could have affected the results, such as participants’ location or socioeconomic status, were simply unavailable to the researchers. 

“Using data from the public domain to be able to build transcriptional signatures or features in the immune system that can be detected pre-vaccination and can inform about the responses to the vaccine—and in this case, multiple vaccines—is the right way of doing immunology,” says David Furman, an immunologist at Stanford University School of Medicine who was not involved in the work. “I think that the approach is sound [and] is probably much more translatable than any other approach.” 

“They have done an excellent job in synthesizing the common themes among all vaccine responses,” writes John Belmont, a geneticist at the Baylor College of Medicine who was not involved in the study, in an email to The Scientist. 

How genes affect vaccine-induced immunity 

Thomas Hagan, a former postdoc in Pulendran’s lab and now an immunologist at Cincinnati Children’s hospital, explains that the main goal of the study led by that lab was to determine whether they could predict an individual’s eventual response to a vaccine shortly after it’s delivered. “What we found was that, by and large . . . different types of vaccines” induce similar, conserved genetic pathways, “and typically early on,” he says. 

Most participants, on days one and three after vaccination, showed changes in the expression of genes related to innate immune pathways, including those involved in detecting viruses and spurring on other immune cells, the researchers found. Then, around day seven, cells began churning out mRNA from genes related to the adaptive immune pathway, and particularly genes involved in the cell cycle and those typically expressed by plasma cells and plasmablasts, the cells that make antibodies and are thought to give people short- and long-term immunity.

See “Pfizer Vaccine Induces Immune Structures Key to Lasting Immunity

There were exceptions to this overall pattern, however. “The main interesting thing we saw was that there was a lot of difference” in how quickly the different types of immune response kicked in, says Hagan.

For example, the yellow fever vaccine broke from the usual trend. There’s only one yellow fever vaccine, and the jab is among the most effective vaccines in terms of preventing infection by its target pathogen, explains Pulendran. It also prompted the slowest immune response of all of the vaccines the researchers tested. Some features of the adaptive immune response, such as genes associated with T cells and B cells, kicked in right away. But the innate immune system, which typically kicks in around one to three days after vaccination, didn’t respond until seven days postvaccine. The rest of the adaptive immune response, which typically kicks in after one week, didn’t begin until around day 14 after vaccination. 

Pulendran and Hagan say that this leisurely pace may be related to how long protection from vaccination lasts. “Durability is a topic of great interest these days for all of us,” says Pulendran. “We’re all worried that . . . COVID vaccines are not inducing durable responses [in terms of protection from infection]. But yellow fever induces durable responses that can last for 30 years or 40 years, just after one shot. . . . I think the delayed plasmablast response is giving us a clue” about why.

“Slow and steady wins the race,” suggests Hagan. Although the researchers don’t know why the yellow fever vaccine elicits such a late response, Pulendran says they intend to find out in future studies.

“The big impact is in seeing the differences among vaccines in their [timing] of response. This is to be expected based on the wide variety of vaccine compositions, but it is very important to have actual rigorous meta-analysis to support the conclusions,” says Belmont. 

The next step for the researchers was to ask which signatures in the days and weeks after vaccination were predictive of increased immunity, measured as the quantity of antibody against the target pathogen that was in the blood after about one month. The researchers found that high expression of plasmablast-associated genes, regardless of the time when their expression peaked, predicted a robust antibody response. 

Based on evidence from previous studies together with this new study, “the reality is some people are better protected than others [after vaccination] . . . and it has something to do with these plasmablasts,” says Michael Snyder, a geneticist at Stanford University School of Medicine who was not involved in either study. “The idea of coming up with a signature that can predict that is a really good thing. That’s what they were able to do.”

How genes affect immunity pre-vaccine

The other study dug into the pre-vaccination signatures that correlate with a robust antibody response to a vaccine down the road. The authors compared the gene expression profiles of immune cells pre-vaccination, finding that they could divide study participants into three groups based on their expression of genes associated with the inflammatory response, metabolism, and cell proliferation. 

One group, which the team termed the high inflammatory endotype, had high expression levels for genes that are hallmarks of inflammation. These included the interferon pathway, the nuclear factor-kappa B pathway and its target cytokines, and the IL-6 signaling pathway. The medium and low inflammatory endotypes had lower levels of these genes. 

Across all 13 vaccines in the dataset, study participants with the high inflammatory endotype had significantly larger vaccine-induced antibody responses than those with other endotypes. The authors write in the paper that it’s possible that immune cells expressing these inflammation-related genes are presensitized, or “primed,” to kickstart an immune response. 

“I think it’s intriguing that what they found is that pre-vaccination, inflammation is actually a positive predictor of the vaccine response,” says Furman. “That’s counterintuitive to me,” he adds, because other studies have shown that baseline levels of inflammation can lead to a weaker immune response. 

Sékaly counters that “if you really want to have a very good immune response, you need to have inflammation. There’s good inflammation and there’s bad inflammation.” He adds that vaccines ideally trigger “good” inflammation, which leads to a protective immune response. 

Furman notes that the authors only analyzed the immune responses of healthy individuals. He says he would have liked to have seen an analysis of whether different subgroups of participants, such as the elderly or the immunocompromised, had unique pre-vaccine transcriptional signatures of immunity. 

The team used a machine learning approach to predict whether an individual would have a high or low response to vaccination based on their pre-vaccine gene expression. They defined ‘high’ and ‘low’ as the top and bottom 30 percent of responders based on mean antibody levels in each study (the middle 40 percent of responders were excluded from the analysis). But the method is limited in how accurately it can predict someone’s response to a vaccine, the researchers say. “At our best, we do two-thirds of our predictions correctly,” says Slim Fourati, a postdoc in Sékaly’s lab. For the other one-third of people, gene expression doesn’t seem to predict immune response. 

The authors say they hope that the study will help researchers zero in on potential therapeutic targets and ways to boost immunity in people who otherwise would have a weak response to vaccines, or to design vaccines that work better for certain endotypes. “If we are able to predict, before vaccination, if someone is going to respond or not to a vaccine, we can try to adapt the vaccine . . . to the specific host,” says Sékaly. 

Similarly, Pulendran says that the findings from the study he led could someday help researchers make diagnostic technologies such as a “vaccine chip” to determine whether someone will develop a protective immune responses to any vaccine, “instead of having to wait a month or longer,” he says. Weak responders could then perhaps be given a tailored shot or treatment.  

The researchers stress that all of the data is now online for anyone to use and say they hope that scientists will take advantage of the large resource. “We think this is going to be an important resource and a benchmark for the future as people develop vaccines against emerging diseases and pandemics,” Pulendran says. 

ADVERTISEMENT