One flu virus can hamper the immune response to another

Prior exposure to one strain of influenza virus may weaken children's ability to mount an effective antibody response against subsequent exposure to a different flu strain, according to a study led by Weill Cornell Medicine investigators.

The analysis of the pediatric response to H3N2 and H1N1 influenza A viruses, two of the most common causes of flu, provides insight into the concept of "immune imprinting," and supports the idea that childhood vaccination, if properly designed, may be able to mitigate its adverse effects.

Immune imprinting, first described in 1960, refers to a long-term bias or skew in the immune response that a first viral exposure can introduce, with potentially serious impairment of subsequent responses to closely related viruses. Imprinting has been difficult to research and is far from being well understood.

However, in the study, published March 11 in Nature, the researchers were able to analyze antibodies from participants, including children 2 to 6 years old who had their sequential first exposures to the two types of influenza A viruses circulating in humans. They determined that following an initial H3N2 exposure, antibodies generated that could bind both H3N2 and H1N1 were ineffective at stopping most H1N1 strains.

Structural analyses and other experiments linked this imprinting effect to a key site on the viral antigen that is similar on H3N2 and H1N1. But by first shaping the immune response to the H3N2 virus, the children became unable to properly adapt to this site on the H1N1 virus. The researchers also found evidence that simultaneous vaccination in infancy against H3N2 and H1N1 may eliminate this imprinting problem.

"These findings are a reminder that the effectiveness of a vaccine against a given virus may be impacted by a past exposure to a related virus, which can make vaccine design a lot tougher," said senior author Patrick Wilson, the Anne E. Dyson Professor of Pediatric Research and a member of the Gale and Ira Drukier Institute for Children's Health at Weill Cornell Medicine.

The collaborative study included contributions from many institutions including Scripps Research, St. Jude Children's Research Hospitaland the University of Michigan School of Public Health.

A first exposure to a virus induces an initial, "de novo" B cell response, including secretion of serum antibodies along with the creation of a long-lived pool of memory B cells, which can produce antibodies rapidly if they ever encounter the virus again. The imprinting problem is thought to arise when a related virus triggers this memory response, creating a surge of antibodies that may not be capable of effectively neutralizing the new threat - and may even hinder the immune system's ability to mount a robust de novo response. The problem is thought to be directly linked to the immunity elicited by initial exposures in early childhood.

The new study covered 40 participants, half of whom were young children with confirmed first exposures to H3N2 then H1N1, or vice versa. Participants came from two larger, clinic-based studies that enrolled children with flu symptoms and analyzed their immune responses - and often those of other household members - for years, to better understand the development of immunity to flu viruses.

"This was a really unique cohort of participants providing us an opportunity to study imprinting of influenza virus through a relatively clean analysis," said first author Jiayi Sun, a postdoctoral associate in pediatrics in the Wilson laboratory.

As expected, the adult participants showed signs of having built up stronger and more versatile immunity through a lifetime of exposures. In some of the children, however, imprinting that impaired immunity was evident.

The children who had encountered the two flu subtypes sequentially over time ended up inducing "cross-reactive" antibodies capable of binding both subtypes at the stalk region of the flu virus's main outer envelope protein hemagglutinin - a site that is almost the same in both subtypes and therefore of major interest to improve future influenza vaccines that might protect against all influenza strains.

In children with H3N2-then-H1N1 exposures, compared to the other groups, these cross-reactive antibodies generally bound much better to the H3N2 stalk than to H1N1's, and were relatively ineffective against the encountered H1N1 strain. Experiments also showed that these H3-imprinted antibodies, binding weakly to the H1N1 stalk region, could block binding by an anti-H1 stalk antibody that more potently neutralized the virus.

The H3N2-imprinted antibodies that appeared after H1N1 exposure were also strikingly ineffective against older H1N1 strains, which may pose a threat in the future. The researchers, using cryo-electron microscopy-based analyses, determined that this loss of activity was likely due to a single amino-acid change differing by only an atomic group in the stalk region in these older H1N1 strains, compared to more recent H1N1 and H3N2 strains. The impact of such a minute molecular difference on the activity of nearly all antibodies to the shared stalk epitope from seven different children was surprising.

"This shows that it doesn't take much to render your memory antibody response ineffective," Wilson said.

B cells from infants vaccinated simultaneously against H1N1 and H3N2 - with a seasonal flu vaccine that typically contains both - showed no sign of this deleterious imprinting.

Currently the American Academy of Pediatrics and other public health authorities recommend the seasonal flu vaccine for all children age 6 months and older.

Many Weill Cornell Medicine physicians and scientists maintain relationships and collaborate with external organizations to foster scientific innovation and provide expert guidance. The institution makes these disclosures public to ensure transparency. For this information, see profile for Dr. Patrick Wilson.

Jim Schnabel is a freelance writer for Weill Cornell Medicine.