&Bullet; physics 14, 72
Researchers use non-equilibrium physics statistical methods to guide the design of vaccines that are effective against many strains of the virus, a holy grail of immunology.
Just over a year ago, physicists around the world shifted their research efforts to infectious diseases to help fight the COVID-19 pandemic. In the United States, soft matter physicists tested a technique to detect SARS-CoV-2, the virus that causes COVID-19. In Switzerland, high-energy physicists tested an inexpensive ventilation technology that could make it easier for those affected to breathe. In the UK, condensed matter physicists donated their personal equipment and protective gear to the country’s hospitals.
Applying the tools of physics to infectious diseases is an old idea, however, as physicists have worked for decades to understand how diseases spread and how viruses develop. Arup Chakraborty from the Massachusetts Institute of Technology is one of those physicists. He and his postdoc Raman Ganti now present the results of their latest project on the use of statistical-mechanical methods to predict the properties of the optimal antigens for “universal” vaccines for the rapid mutation of viruses such as influenza  . Their findings should also be applicable to the development of vaccines with broad coverage for a range of viruses from HIV to SARS-CoV-2.
“Vaccinations saved more lives than any other medical procedure,” says Chakraborty. “Even so, there are no effective vaccines against highly variable pathogens. That is why we need a seasonal flu vaccination. Now that COVID-19 is mutating, we might need repeated vaccinations for this virus as well. “
Every year medical professionals inject hundreds of millions of flu vaccines into the arms, legs and buttocks of people around the world. These vaccines contain an antigen from a weakened or inactivated strain of the flu virus that causes the body to make antibodies. These antibodies target so-called spike proteins that protrude from the viruses and are used by the viruses to enter cells.
Different antibodies target different regions of the spike protein, so the body goes through an evolutionary process that involves so-called B cells to choose which antibodies to develop. The most commonly chosen influenza virus antibodies target the regions of spike proteins that mutate rapidly. This means that as soon as a new strain of flu develops, the virus can evade the body’s defenses and a new vaccine must be developed. However, there are regions of spike proteins that rarely vary.
Immunologists want to develop a vaccine that will trigger the production of antibodies that target these unchanging regions. These antibodies, known as largely neutralizing antibodies, have been isolated from some people who are naturally infected with influenza or HIV. However, they rarely occur and are not seen with these vaccines, says Chakraborty. “The question is, how can we vaccinate so that these rare antibodies become dominant?”
To answer this question, Chakraborty and Ganti turned to methods from statistical non-equilibrium physics and created a model that predicts the type of antibody produced. They opted for a two-vaccination protocol, a common strategy where the first shot is followed by a booster shot shortly afterwards. In the model, each of the vaccines contains a different composition of virus antigens – basically different strains of the inactivated pathogen. The body’s B cells respond to the antigens by rearranging their population in an evolutionary process that resembles a disturbed system reaching equilibrium. The model controls how disruptive each antigen is based on how different it is from the antigens that the B cells “know” from previous exposures.
To interpret the results, the duo applied a statistics-based learning theory that is widely used in educational research and machine learning. “The development of antibodies is a learning process for the body, so we thought that the theory could provide interesting insights,” says Ganti.
As part of this theory, the duo found that largely neutralizing antibodies can dominate under certain learning conditions. First, the antigen in the original vaccine must not be too similar or too different from what the body has already encountered. In other words, the level of learning difficulty for the B cells cannot be too easy or too severe or the final antibodies that the cells develop will not largely neutralize. “The right antigen for the first vaccination is crucial,” says Ganti. Second, the level of learning for the booster shot must be higher than the first. (Again, there’s a sweet spot for how high.)
If these learning steps are correctly optimized, the model shows that largely neutralizing antibodies will dominate. “The results show that if we learn properly, we can control antibody development so that an otherwise rare pathway becomes much more likely,” says Chakraborti. The duo are now working to translate this learning strategy into an optimal genetic sequence that will allow researchers working on universal vaccines to select specific antigens for the two vaccinations.
Arvind Murugan of the University of Chicago says the new results “suggest a more systematic and principled way to identify optimal vaccination protocols.” Murugan is particularly fond of the study’s use of learning theory. “Basically, developing a largely neutralizing antibody is like learning the right lessons from your history: you want to focus on aspects of your previous experience that are relevant to future challenges, rather than idiosyncratic details that will never be relevant again . ” he says.
– Katherine Wright
Katherine Wright is the assistant editor of physics.
- RS Ganti and AK Chakraborty, “Mechanisms on which vaccination protocols are based and which can optimally produce largely neutralizing antibodies against highly variable pathogens”. Phys. Rev. E.103052408 (2021).