Part 2: This is the second of a three-part series on how Stanford Medicine researchers are designing vaccines that protect people from not merely individual viral strains but broad ranges of them. The ultimate goal: a vaccine with coverage so broad it can protect against viruses never before encountered.
The series opener focused on why having vaccines that cover not just one strain of a single virus, but many, could be an invaluable advance. Although Part 1 focused on the COVID-19-causing SARS-CoV-2 as a textbook example, that strain-spewing microbe is just one of many viruses to worry about.
Take, for instance, the old steady: influenza.
Like SARS-CoV-2, the influenza virus is studded with molecular hooks that it uses to latch onto vulnerable cells in our airways and lungs. Analogous to the spike protein, the influenza virus's hook-like molecule, called hemagglutinin, is the principal antigen in the influenza vaccine.
The hemagglutinin molecule's head is quite immunogenic. Once antibodies have been generated, their binding to the hemagglutinin head can impede (or, ideally, totally block) the virus's cell-invasion process.
The influenza virus's hemagglutinin is, alas, highly subject to mutation, reliably spinning off new strains each year. Each year's vaccine version is based on vaccinologists' best-guess prediction of which four influenza strains are deemed most likely to be circulating in the U.S. during the winter flu season.
Even when that prediction is dead-on, said Mark Davis, PhD, professor of microbiology and immunology and the Burt and Marion Avery Family Professor, a large fraction of vaccinated people fail to develop antibodies to one or more of the strains represented in the vaccine. Which of those strains that turns out to be is strongly influenced by genetics, according to an analysis of identical twins conducted by Vamsee Mallajosyula, PhD, a basic science research associate in Davis's lab.
The researchers found this "strain preference," or uneven immune response to different strains, in most of the people they studied -- even infants never exposed to influenza or the vaccine for it.
Davis' group has found a way to trick our immune system into paying attention to all four strains represented in the vaccine.
"Our work is an outlier, a different approach," he said.
B cells, our body's antibody factories, double as antigen presenting cells. But they're ultra-picky about what they pay attention to. Just as an individual B cell will produce only one species of antibody fitting a mere one or very few antigenic shapes, that B cell recognizes and ingests only material containing that same antigenic feature or ultra-close look-alikes.
Depending on genetic luck of the draw, a person's immune repertoire may not include any, or not many, B cells capable of displaying or producing antibodies to fit this or that strain represented in the vaccine. In that case, the person won't develop a sufficient immune response to the strain (or strains).
To overcome strain preference, Davis, Mallajosyula and their colleagues designed a vaccine in which the four hemagglutinin varieties are chemically stapled together. As a result, any B cell that recognizes one or another of these four hemagglutinin types gobbles up the entire matrix and displays bits of all four on its surface. Other immune cells called helper T cells assist in activating B cells, but they are just as finicky as B cells are about what specific antigens they respond to, activating only B cells displaying those antigens. When B cells with a preference for one strain nevertheless wind up displaying bits of hemagglutinin from all four strains, T cells are more likely to stumble on that antigen they love to hate -- and to activate the B cell sporting it.
Davis and colleagues tested their four-antigen vaccine matrix by putting it into cultures containing human tonsil organoids -- living lymph tissue originating from tonsils extracted from tonsillitis patients and then disaggregated. In a laboratory dish, the tissue spontaneously reconstitutes itself as small spherical tonsil "mini-me's" that act just like lymph nodes -- ideal environments for antibody manufacturing.
Sure enough, B cells in these tonsil organoids recognizing one or another of the four conjoined hemagglutinin molecules swallowed the whole matrix and displayed bits of all four varieties, thus recruiting far more helper T cells to kick-start their activation. This resulted in solid antibody responses to all four influenza strains.
Davis thinks using this kind of construct could substantially elevate the efficacy of the influenza vaccine, which is now roughly 20% to 80% effective depending on how well the pre-season guesswork has predicted the dominant strains in circulation when the season begins.
Keep your heads down
The influenza virus's hemagglutinin molecule mutates frequently almost entirely at its "head," the part that grabs onto the cell the virus has evolved to invade.
The hemagglutinin molecule's nether region, or "stem," doesn't mutate much at all, said Peter Kim, PhD, the Virginia and D.K. Ludwig Professor of Biochemistry. Antibodies that bind to stem portions could, in principle, incapacitate many influenza strains no matter how much the head mutated. On the other hand, the hemagglutinin stem normally isn't immunogenic: It doesn't generate much production of antibodies to itself.
Duo Xu, PhD, a postdoctoral scholar in Kim's lab, has found a way to focus the immune system's attention on the hemagglutinin molecule's unchanging but unseen stem. Xu clumped together hemagglutinin threesomes at their heads, obscuring those heads from immune view. Instead of the heads, the stems stuck out. In this way, he transformed the stem into a stretch of bull's-eyes for immune cells. That boosted the production of antibodies targeting an unprecedented range of influenza strains -- "broader binding than has been observed before," Kim said.
Part 3: Going beyond B cells in the search for a more multi-targeted vaccine
Part 1: The hunt for a vaccine that fends off not just a single viral strain, but a multitude
Photo: Numstocker
