How have influenza vaccine designs changed over time?

This answer examines universal influenza vaccine design as discussed by the paper the question references,

A universal flu vaccine should be able to protect against all flu strains and subtypes including those that cause seasonal as well as pandemic flu. Doing so would render unnecessary annual flu vaccines whose several shortcomings include distinct disadvantages in the face of potential epidemic or pandemic flu,

  • Need to be reformulated each year.
  • Are based on unreliable early snapshots of flu strains circulating any given year.
  • Often end up missing the mark in matching the actual circulating strain(s).
  • Take months to make ready.
  • Can’t be reformulated on the fly late in the season to improve their match to circulating flu strains.
  • Tend to be less effective in the elderly, one of the groups most susceptible to flu-related morbidity and mortality.

The paper discusses flu vaccines from a structural biology perspective, a fashionable notion for quite some years now. Often called Reverse Vaccinology (RV), though it’s more accurately Structure-based Reverse Vaccinology, the basic idea begins with an analysis of the structure of those antigen-antibody complexes where an antibody is found to be broadly neutralizing.

Broadly neutralizing means a given antibody is able to effectively bind and neutralize a broad range of structurally related epitopes (bits of antigens that bind paratopes, specific counterparts of antibodies), typically from related strains and subtypes of a given pathogen such as flu. In the form of a broad adaptive immune response, bnAb has thus come to represent a holy grail of sorts in diseases such as flu, HIV, HCV that currently lack effective, long-lasting vaccines.

Essential idea of RV is technological, not immunological. Figure out the sequences of antigenic epitopes that strongly bind bnAb, synthesize such sequences and use them as vaccines with some masala (nanoparticle, etc.) thrown in for good measure. Similar bnAb will ensue against a wide variety of strains and subtypes, and render annual reformulations of seasonal flu vaccines unnecessary or so the idea goes.

RV’s chief assumptions are

  • Antigenic epitope binds to cell-surface B cell receptor (BCR) similarly as it does the secreted bnAb.
  • bnAbs bind the same antigenic epitopes as their initial antibody precursors.
  • Antigenicity equals immunogenicity.

While this notion sounds simple and straightforward, immunology, unlike technology, is anything but simple as becomes evident from scrutinizing problems with these assumptions, which are not one but all too many.

  • How free antigen binds free antibody is not the same as antigenic epitope binding BCR, i.e., surface antibody.
    • Intact molecules are more complex with their antigenic and immunogenic components often located in different parts.
    • How cell-surface BCR and its secreted counterpart (antibody) bind may not be exactly the same with the the cell membrane’s lipid bilayer serving to stabilize binding in the former.
  • Antibodies are inherently polyspecific, capable of binding more than one antigenic epitope. What’s found bound in an antigen-antibody complex may not be the only molecule that that antibody could bind nor the one it could bind most strongly or effectively.
  • Antigenicity does not equal immunogenicity. Rather, akin to the proverbial iceberg, the vast glacier of immunogenicity lies below the tip of antigenicity. While decoding the exact sequence of an antigenic epitope that binds a neutralizing antibody is easier than ever these days, that alone does not recapitulate the myriad biological activities that occur in tandem in nature to get an immune response to that point in the first place.
    • Antigenicity pertains to the physical lock and key type interaction between an antigen and a somatically rearranged receptor expressed by a B or T cell that allows them to bind some antigenic epitopes and not others while immunogenicity encapsulates the far more complex sequence of events that comprise an immune response.
    • Various molecules expressed by a pathogen trigger various types of germline encoded receptors in innate immune cells as well as somatically rearranged germline receptors of adaptive immune T and B cells.
    • Linked recognition of certain pieces (epitopes) of such molecules by both CD4 T helper cells as well as B cells ensures the former provide ‘help’ to the latter critical in ensuring their antibodies against such epitopes improve in their binding strength (affinity) over time.
    • In the process, even what a B cell binds after it undergoes such changes during the course of an immune response can be different from what it bound initially (1), a consequence of the process called Somatic hypermutation – Wikipedia that’s unique to B cells. Also a sobering reminder that adaptive immune system targeting of antigens is dynamic, not static.

Simply decoding the sequence of what is found bound to a bnAb does not help recapitulate any of these processes and to focus vaccine efforts on that alone is akin to proposing the tip of the iceberg informs everything about the iceberg from its size to shape and depth. Zooming out to examine not just the Z but the A-to-Z of how an immune response gets to a bnAb shows at least for viruses that brute technological prowess can’t supplant how the immune system operates, at least not yet.

What emerges as a bnAb is the outcome of complex cellular and molecular processes that endow antigen-specific responding B cells with the capacity to secrete antibodies that become progressively more effective and capable of neutralizing the antigens they bind. T cell help for B cells is a key component of that process that RV indefensibly chooses to ignore. Even most basic questions pertaining to a given bnAb remain woefully ignored and unexplored for the most part.

  • What are the corresponding T cell specificities or what are the targets of T cells that help B cells make bnAbs?
  • What are the corresponding T cell effector classes (cytokine responses)?
  • What about non-neutralizing but yet inhibitory antibody responses and the T helper cells that support them? How to maximize their potential?
  • What about ineffective antibody responses and the T helper cells that support them? How to prevent them from taking over and/or how to neutralize them?

In a manner that simply beggars comprehension, bnAb analysis is typically undertaken in a sui generis fashion as though what we now understand of how the immune system works simply doesn’t exist or doesn’t matter.

These issues go a long way in explaining why each time it’s been empirically tested for HIV, RV abjectly failed time and again (2, 3). HCV provides an example of how perceived utility of bnAbs are also far from generalizable since anti-HCV bnAbs have been found to be not protective (4).

While RV succeeded in the case of bacterial meningococcus vaccines, luck rather than design may have played an outsize role in this case of encapsulated extracellular bacteria that are mainly controlled by neutralizing antibodies. Such biological simplicity and knowledge is sorely lacking with viral diseases such as flu, HIV, HCV. Two other major problems further highlight the inadequacy of reverse-engineered vaccines with respect to such viral targets.

  • Rote reliance on experimental animal models that poorly recapitulate human disease and are thus poorly predictive of human immune responses against such pathogens.
  • Lack of knowledge as to which types of immune responses are necessary and sufficient in terms of effectiveness (correlates of protection).


1. Xiao, Xiaodong, et al. “Germline-like predecessors of broadly neutralizing antibodies lack measurable binding to HIV-1 envelope glycoproteins: implications for evasion of immune responses and design of vaccine immunogens.” Biochemical and biophysical research communications 390.3 (2009): 404-409.…

2. Van Regenmortel, Marc HV. “Structure-based reverse vaccinology failed in the case of HIV because it disregarded accepted immunological theory.” International journal of molecular sciences 17.9 (2016): 1591.…

3. Pollara, Justin, David Easterhoff, and Genevieve G. Fouda. “Lessons learned from human HIV vaccine trials.” Current Opinion in HIV and AIDS 12.3 (2017): 216.…

4. Giang, Erick, et al. “Human broadly neutralizing antibodies to the envelope glycoprotein complex of hepatitis C virus.” Proceedings of the National Academy of Sciences 109.16 (2012): 6205-6210.…