Jan 3, 2007 (CIDRAP News) – California scientists report that their analysis of the medical literature has yielded data on more than 600 molecular components of influenza A viruses that trigger immune responses, findings they hope will spur the search for vaccines offering protection against multiple flu strains.
Scientists at the La Jolla Institute for Allergy and Immunology (LIAI) scanned more than 2,000 scientific articles in a hunt for data on influenza A molecular structures that interact with either of two major components of the immune system: T cells and antibody-producing B cells.
Writing in the Jan 2 Proceedings of the National Academy of Sciences, the researchers report that they found information on 602 such structures, called epitopes, from 13 different influenza A subtypes. These included one particular epitope that is shared by several human flu subtypes and the H5N1 avian flu virus.
Flu vaccines now in use target two surface proteins, hemagglutinin and neuraminidase. Because these proteins constantly mutate, vaccines must be changed every year to match the strains expected to be circulating during the flu season. But scientists have long hoped to develop a vaccine that would target a viral protein that is "conserved," or essentially the same, in different strains. Epitopes that different strains have in common could, scientists believe, be used to make a vaccine that would protect against multiple strains and could be used for years.
"If we can find shared epitopes, it may be possible to develop an influenza vaccine with greater cross-protection for many different viruses," Alessandro Sette, PhD, senior author of the study, commented in an LIAI news release.
The authors used the Immune Epitope Database (IEDB), described as the world's largest repository of data on immune responses to infectious agents, to search the literature for influenza A epitopes. The La Jolla institute developed the IEDB with a $25 million contract awarded by the National Institute of Allergy and Infectious Diseases (NIAID) in 2004. Sette said the IEDB collects all known antibody and T cell epitope information in one place and is available to scientists around the world (see link below).
The researchers found 2,063 articles related to influenza A epitopes, of which 429 were deemed worthy of detailed examination. This led to the cataloging of 412 T cell epitopes and 190 antibody epitopes from 13 viral subtypes and 58 different strains. The analysis yielded important data but also revealed important information gaps, the researchers write.
The relatively low number of antibody epitopes was surprising, the authors write, given that antibody titers are the only accepted measure of protection from flu.
Only two H5N1 avian flu epitopes were found, both from a 2004 Vietnam strain of the virus. The lack of H5N1 epitope data is not surprising, given the recent emergence of the virus and the special biosecurity measures required for studying it, the researchers say.
Antibody epitopes were identified from only 5 of the 10 viral proteins—most of them from hemagglutinin, neuraminidase, and M2—whereas T cell epitopes from all 10 proteins were identified. Only one antibody epitope—versus 160 T cell epitopes—was identified by studying human samples (rather than animal samples). The authors comment that interpreting epitope data from human samples is more complex because people, unlike lab animals, typically have been exposed to many flu strains over many years.
Using an analysis tool developed as part of the IEDB, the authors found that a higher percentage of T cell epitopes than of antibody epitopes were shared by multiple viral strains. About 11% of T cell epitopes were 100% identical in human and avian strains, while 30% of them were 90% identical, and 50% were 80% identical. In contrast, only 2.7% of antibody epitopes were 100% identical, and less than 11% were found to be 80% identical.
"In general, the results suggest that significant levels of interstrain cross-reactivity are likely for T cell epitopes, but much less so for Ab [antibody] epitopes," the report says.
Because not all antibody and T cell responses are protective, the researchers sifted their data for epitopes associated with protective immune responses in the lab. They found only 9 antibody and 9 T cell epitopes that met their criterion. Most of the protective T cell epitopes are found in both human and avian flu strains, while most of the antibody epitopes are not, they report.
"However, one protective Ab epitope from the M2 protein shows appreciable conservation among the selected human influenza strains and H5N1," the article states. "Because M2 is a relatively conserved protein, identification of protective Ab epitopes derived from this protein, as has been pointed out, holds promise for the future development of a universal influenza epitope-based vaccine."
The authors recommend research to address the gaps their analysis revealed, including more studies on antibody epitopes and efforts to identify more avian flu virus epitopes.
Gregory A. Poland, MD, a flu vaccine expert at the Mayo Clinic in Rochester, Minn., applauded the report, saying it should advance the search for a flu vaccine that targets a conserved viral component and offers protection against multiple strains. He directs the Mayo Vaccine Research Group and Program in Translational Immunovirology.
"If you find a wholly conserved piece of the virus that doesn't vary much between strains and provides cross protection between [H5N1] clade 1 and clade 2 viruses, that could be very exciting and could be the basis for developing a subunit vaccine," Poland told CIDRAP News. "You could make it quickly and in huge quantity."
He also said the report could generate some controversy, given the division of opinion among vaccine experts on whether flu vaccines should be made from whole viruses or from viral subunits. Existing evidence shows that whole-virus vaccines are more immunogenic, but there is very little experience with subunit vaccines, he said.
"The vaccine world is divided into whole-virus and subunit type people," Poland said. "I happen to believe there's a lot of merit to the subunit and peptide approach."
Poland also praised the LIAI researchers for collecting all the influenza A epitope data into a single database and making it possible to do types of analyses that couldn't be done before. "It does expand the database and advances the knowledge," he said.
The NIAID, which supported the LIAI research, hailed it in a news release: "The study should help scientists who are designing new vaccines, diagnostics and immune-based therapies against seasonal and pandemic influenza because it reveals in molecular detail exactly where the immune system focuses on the viruses. . . . Information on shared protective epitopes is important for developing influenza vaccines that can provide broad protection against multiple strains of the virus."
However, the agency cautioned that the identification of conserved epitopes doesn’t necessarily mean that broadly protective vaccines are possible. "What is less clear from the analysis is how cross-reactive an immune response would be to most of these conserved epitopes," the news release states. "Further analysis may assist scientists in identifying vaccine targets that might offer broader protection and in predicting how effective a new vaccine will be."
Bui H-H, Peters B, Assarsson E, et al. Ab and T cell epitopes of influenza A virus, knowledge and opportunities. Proc Natl Acad Sci 2007;104(1):246-51 [Abstract]
Immune Epitope Database and Analysis Resource homepage
Jan 1 LIAI news release via EurekAlert
Jan 3 NIAID news release
Aug 25, 2005, CIDRAP News story "Acambis hopes to build a flu vaccine that lasts