Creating a versatile vaccine to combat Covid-19 in many ways | MIT News

According to ancient lore, one of Hercules’ 12 deeds was to defeat a nine-headed monster called the hydra. The challenge was that if Hercules used his sword to cut off one of the monster’s heads, he would grow two in its place. Therefore, he needed an additional weapon, a torch, to defeat his enemies.

There are parallels between this legend and the three-year battle with SARS-Cov-2, the virus that causes Covid-19. Each time a scientist thought he had put down one lineage of viruses (alpha, beta, delta, ohmicon, etc.), another variant or subvariant emerged some time later.

For this reason, researchers at MIT and other institutions are preparing new strategies against viruses. This is a new vaccine that could potentially fight all subspecies of the disease, unlike those currently in use, with a property called ‘pan-dispersion’. Avoids the need for another booster shot each time a new strain is circulated.and paper Published in a magazine today The forefront of immunologyteam A report on experiments in mice demonstrating the effectiveness of vaccines in preventing death from Covid-19 infection.

Viral vaccines usually work by exposing the immune system to small fragments of the virus. This allows you to create learned reactions that will protect people later when exposed to the real virus. The premise of standard Covid-19 vaccines like those produced by Moderna and Pfizer is to activate parts of the immune system that release neutralizing antibodies. By providing cells with instructions (in the form of mRNA molecules) to make the spike protein, a protein present on the surface of the Covid-19 virus whose presence can trigger an immune response. do this. “The problem with this approach is that the target keeps changing” — the spike protein itself can vary between different virus strains — “and that can render the vaccine ineffective,” says MIT Electric. says David Gifford, professor of engineering and computer science and biology.engineering and co-author frontier paper.

Therefore, he and his colleagues took a different approach and chose a different target for the vaccine. It activates the part of the immune system that unleashes “killer” T cells that attack virus-infected cells. This kind of vaccine cannot prevent people from contracting Covid-19, but it can prevent serious illness and death.

A key innovation made by this group, which includes researchers from MIT, the University of Texas, Boston University, Tufts University, Massachusetts General Hospital, and Acuitas Therapeutics, was the introduction of machine learning techniques into the vaccine design process. A key aspect of that process involves deciding which parts of SARS-Cov-2, which peptides (chains of amino acids, which are the building blocks of proteins), will be included in the vaccine. This requires sifting through the thousands of peptides in the virus and selecting only 30 or so that need to be incorporated.

But that decision has to take into account the so-called HLA molecules, protein fragments on the surface of cells that act as “signs” and tell what’s going on inside other cells. Tells immune cells (which lack X-ray vision). there is.

Machine learning algorithms were used to solve a series of complex “optimization problems,” said an affiliate of the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) and a PhD in electrical engineering and computer science at MIT. Course student Brandon Carter points out. First author of a new paper. The overarching goal is to select peptides that are present or ‘conserved’ in all variants of the virus. But these peptides also need to associate with his HLA molecules, which are likely to be displayed, so that they can alert the immune system. “We want as many people as possible to have this happen to maximize vaccine coverage,” says Carter. In addition, he added, he hopes that each individual will be covered multiple times by the vaccine. Achieving these various objectives is a task that can be greatly facilitated by machine learning tools.

While that touches on the theoretical end of the project, the latest results come from experiments conducted by collaborators at the University of Texas Medical School in Galveston, which show potent viral cytotoxicity in vaccinated mice. The mice in this experiment did not die, but were “humanized.” That means there were HLA molecules found in human cells. “This study proves in a living system, in real mice, that a vaccine devised using machine learning can provide protection against the Covid virus.” It has been characterized as the first evidence that a vaccine formulated in

Paul Offit, professor of pediatrics in the Department of Infectious Diseases at Children’s Hospital of Philadelphia, finds the results encouraging. “Many people are wondering what approaches will be used to make Covid-19 vaccines in the future,” says Offit. Cellular responses will be an important step forward in the next generation of vaccines. “

More animal and, ultimately, human studies are needed before this work can usher in a “next-generation vaccine.” The fact that 24% of his lung cells in vaccinated mice were his T cells “showed that the mouse’s immune system was ready to fight viral infection,” he said. said. But he warns that you have to be careful not to have too strong an immune response so as not to damage your lungs.

I have many other questions. Should T-cell vaccines be used in place of or in combination with standard spike protein vaccines? While it may be possible to enhance existing vaccines by including a T-cell component, Gifford “One part of the vaccine may mask the other, so putting the two together may not be strictly additive.

Nevertheless, he and his colleagues believe that their T-cell vaccine could help immunocompromised individuals who may not be able to produce neutralizing antibodies and thus benefit from conventional Covid vaccines. Their vaccine may also reduce suffering from the ‘long Covid’ for those who continue to carry reservoirs of the virus after initial infection.

The mechanism behind current influenza vaccines, similar to current Covid-19 vaccines, is to induce neutralizing antibodies, but these vaccines do not always work against different influenza strains. . Carter sees potential in his flu vaccine based on T-cell responses.

The method they are developing is not limited to Covid-19 or influenza, and may one day be applied to cancer. Designed to maximize immune protection between individuals, it could be a key asset in the fight against cancer. “It’s not within the scope of current research, but it could be the subject of future research,” he says.

Other MIT contributors were Ge Liu and Alexander Dimitrakakis. This work was partially supported by Schmidt Futures and his C3.ai grant to David Gifford.