Our current influenza vaccines are strain-specific and often ineffective against new variants. The influenza virus mutates rapidly which means we have to constantly change and tweak flu vaccines. A new study investigated a novel approach to developing a universal influenza vaccine. They used something called the Computationally Optimized Broadly Reactive Antigen (COBRA) to find new ways to fight the influenza virus. But can artificial intelligence (AI) truly crack the code for a lifetime flu vaccine?
Seasonal influenza plagues humanity with persistent outbreaks, claiming up to 650,000 lives annually. Its rapid mutation poses great problems so you have to take a vaccine every year just to keep up with it. Influenza also tends to pop up unexpectedly and jump from species to species. Just look at the current ongoing avian flu epidemic and how it’s already affecting cats and even people.
“Avian influenza viruses are of particular concern because they can infect multiple species and lead to unpredictable and severe disease. There is an urgent need for a universal influenza vaccine that provides protection against all influenza strains,” write the authors of the new study, led by Xiaojian Zhang from Georgia University. To tackle it more effectively, we need a universal flu vaccine.
For a universal flu vaccine, look at the proteins
To develop a new type of vaccine, the researchers look at how influenza infects cells. The virus attaches itself to receptors on the cell surface using a protein called hemagglutinin (HA). To a lesser extent, the virus can also use another protein called neuraminidase (NA). Then, once it attaches itself, the virus starts to take over the cell, hijacking its machinery to replicate and produce more viral proteins, infecting more cells, and so on.
Current vaccines work by stimulating the immune system to recognize and fight the virus before it can cause illness. They “teach” your body to identify viruses as foreign, triggering the production of antibodies specifically targeting the virus’ HA protein. However, this approach is strain-specific. It has limited efficacy against unexpected variants or antigenic drift and shift.
Antigenic drift refers to small, gradual mutations in the influenza virus’s surface proteins (like hemagglutinin and neuraminidase), allowing it to evade immune recognition over time. Antigenic shift is a more abrupt, significant genetic change that occurs when two different influenza viruses combine to create a novel strain.
This is where COBRA comes in.
COBRA vs Influenza
COBRA is an AI-driven tool that analyzes historical and circulating viral sequences to construct a database of shifts in HA proteins. It uses machine learning to identify conserved viral epitopes (the part that is recognized by the immune system), minimizing the risk of immune escape. Additionally, AI accelerates preclinical testing by predicting immune responses and guiding formulation adjustments.
Using the COBRA methodology, the team created hemagglutinin (HA) antigens capable of eliciting broad immunity against diverse influenza strains. They also used adjuvants called cGAMP — molecules that act as an immune system stimulant, enhancing immune responses against infections and in vaccines.
The researchers tested their approach on mice. The formulation with the adjuvants induced very high antibody activity and HA-inhibition. Vaccinated mice resisted lethal influenza infections, exhibiting minimal weight loss and pulmonary viral works. In other words, the vaccine worked.
Can we bring this to humans?
The COBRA methodology seems to be effective. It created antigens that elicit broad immune responses across multiple influenza strains, which is very encouraging. However, implementing this approach on humans will be challenging.
AI-driven approaches are already transforming vaccine development, as demonstrated during the COVID-19 pandemic with mRNA platforms. The integration of COBRA with mRNA technology could further expedite the path to a universal flu vaccine.
In tests on mice, the vaccines elicited high levels of antibodies capable of neutralizing diverse influenza strains, including H1, H3, and H5 subtypes. However, the mice used in the study were influenza-naive — they had never been exposed to influenza before. Meanwhile, humans have complex immune histories from prior flu exposures and vaccinations, so it’s not clear how our bodies would react.
The study’s findings in mice are promising but require validation in human clinical trials to assess safety and efficacy in diverse populations. It will likely take a couple of years before we have any major update, but the researchers are confident. A universal influenza vaccine may finally be on the horizon.
“The formulation of COBRA vaccines plus cGAMP MPs is a promising universal influenza vaccine that elicits protective immune responses against human seasonal and pre-pandemic strains,” the researchers conclude.
The study “Multi-COBRA hemagglutinin formulated with cGAMP microparticles elicits protective immune responses against influenza viruses” was published in the journal mSphere.
