The South African. The Brazilian. The U.K.’s “Kent.”

They sound like they could be the names of some new hairstyle. But, as most virus trackers know, they are common shorthand for the new strains of SARS-CoV-2, the coronavirus behind the global pandemic. More transmissible, and in the case of the U.K. strain apparently more deadly, the new variants have forced governments around the world to impose tougher travel restrictions and, in some cases, new lockdowns.

The new variants also pose a problem for the first crop of vaccines. That’s because almost all the vaccines approved so far target the coronavirus’s spike protein. Mutations in this protein can reduce the vaccines’ effectiveness, potentially even negating any immunity.

So far, the solution vaccine makers and governments have proposed is to begin preparing updated versions of the existing vaccines that will prompt the immune system to make antibodies to the modified spike protein found in the new variants.

But if the virus keeps mutating, the world may find it is stuck in a perpetual game of cat and mouse, always trying to catch up with the latest strains of the virus, with a large portion of the world requiring an annual booster vaccination. This is essentially what happens with the flu virus now. And, as with the flu virus, there is a constant risk that researchers will misjudge and fail to spot an emerging and fast-spreading variant that will once again put many people at risk of hospitalization or death.

Might there be another way?

Some scientists think there is: either using more traditional vaccine technology that exposes people to the real virus and all of its proteins, or using new messenger RNA technology to create a universal SARS-CoV-2 vaccine that would be effective against all current and future strains.

Why the variants are so worrisome

First, a little more background on the current situation: While the U.K. “Kent” strain, which is formally designated B.1.1.7, incorporates changes that make it easier to spread and may make it more deadly, the spike protein is not significantly changed, and the approved vaccines work well against it.

But mutations associated with the South African and Brazilian variants of the virus, formally called B.1.351 and B.1.1.248, respectively, render existing vaccines less effective. An analysis of data from the South African clinical trial of the University of Oxford and pharmaceutical company AstraZeneca vaccine showed that it could not prevent mild to moderate illness in those infected with the B.1.351 strain, although the company says it believes the vaccine probably still protects against severe COVID-19. Lab tests using blood samples from those vaccinated with Pfizer’s and Moderna’s messenger RNA-based vaccines also showed that higher antibody levels were needed to defeat the mutant strain than the original virus. But Moderna says it is confident its vaccine produces enough antibodies that it will still protect against moderate or severe disease.

Worryingly, scientists in the U.K. have now discovered a version of the B.1.1.7 “Kent” virus that also incorporates the same spike protein mutation, known as E484K, that the South African and Brazilian strains exhibit. This could have happened if a single person was infected with multiple strains of the virus, which then had a chance to mix their genetic material; or it could be that same mutation occurred spontaneously more than once.

Most of the COVID vaccines that have been approved for use so far were created with relatively new techniques: messenger RNA (mRNA) or modified adenovirus vectors. In both cases, the idea is to instruct human cells to produce one of the coronavirus’s proteins so that it prompts an immune response. The body then produces antibodies that can attach themselves to that protein and disable it. It is also hoped that other parts of the immune system—such as T-cells, which can kill infected cells—learn to recognize the protein as a sign of a foreign invader and kill those cells.

These technologies have some big advantages over traditional vaccine-making methods: They have excellent safety profiles, so the researchers working on the vaccines were reasonably sure they would not cause severe side effects—something which has been borne out in subsequent human clinical trials. The other good thing about them is that they can be tailored to a new virus very quickly, provided that virus has had its genome sequenced and there is an obvious protein to target, as was the case with SARS-CoV-2. The entire reason we have multiple vaccines in millions of people today, less than a year after the World Health Organization declared a pandemic, is largely because of these newer methods for creating vaccines.

The single-protein problem

But one downside of the way these techniques have been used to combat SARS-CoV-2 so far is that they instruct the cells to make a single virus protein. As a result, they are always going to be vulnerable to mutations in that particular protein. SARS-CoV-2 has four main structural protein groups: The S protein, or spike protein, is the best known. But it also has a nucleocapsid or N protein, a membrane or M protein, and an envelope or E protein. It might be possible to create vaccines that prompt an immune response to some or even all of these.

Two traditional vaccine-making techniques expose the body to all of these proteins because they use the actual virus itself. In one method, a living virus is “attenuated,” or weakened, by growing it in a way that makes it hard for the virus to rapidly reproduce. In another, the virus is “inactivated,” or killed using a chemical treatment, and then either administered whole, or sometimes broken into pieces. This is potentially safer than giving someone a living virus vaccine, which have a nasty habit of evolving back into dangerous pathogens.

No one is considering making a live SARS-CoV-2 vaccine, but several companies are working on inactivated ones. The Chinese company Sinovac’s vaccine uses the inactivated virus, and it has already been given to hundreds of thousands of people in China and in places like Brazil. The company says it is effective against the South African variant, but it has not published the data to support that claim. Meanwhile, clinical trials of the Sinovac vaccine in Brazil have shown it is 100% effective at preventing severe disease but may be only slightly more than 50% effective at preventing very mild illness.

Multivalent vaccines

French company Valneva, meanwhile, is working on an inactivated vaccine using the whole virus that might have advantages over those already approved, Thomas Lingelbach, the company’s chief executive officer, says. Because it uses the complete virus, the Valneva vaccine enables the immune system to potentially form a response to all possible epitopes—a term for the portions of the virus’s proteins that the immune system can recognize. Valneva also combines the inactivated virus with an adjuvant, a chemical substance that boosts the body’s immune response. What’s more, Valneva has experience producing multivalent vaccines—those that incorporate multiple virus strains in a single shot—and it could potentially produce one for SARS-CoV-2 too.

Lingelbach calls his company’s efforts the “third wave” of COVID-19 vaccine candidates. He believes they could have a multivalent version of Valneva’s vaccine authorized and available by next spring. (The first wave are those vaccines already approved, and the second wave are those currently in human clinical trials.) The U.K. government has already preordered 100 million doses of Valneva’s vaccine, some of which will be produced at the company’s manufacturing facility in Scotland.

The “universal” vaccine

Another approach may hold out the promise of a universal SARS-CoV-2 vaccine. The idea is to find epitopes that are both capable of eliciting a strong immune response and which are essential for the coronavirus’s reproduction. The idea is that if these proteins are essential for the virus’s life cycle, the virus won’t be able to escape the vaccine through successful mutations.

One company working on this approach is Belgian startup MyNeo. It uses machine learning to try to predict which virus epitopes will trigger the strongest immune response. It then looks for the subset of those epitopes that are found across all coronaviruses, says Cedric Bogaert, the company’s chief executive. He notes there are certain epitopes that are the same across not just all the SARS-CoV-2 variants, but also across the coronaviruses that cause SARS and MERS, as well as those known to infect minks and bats, two species that harbor coronaviruses and from which scientists think future strains may make the leap to humans. These common protein segments are what biologists call “well conserved,” and the speculation is that they don’t change much over time because their function is somehow essential to the virus’s viability.

Of particular interest is the SARS-CoV-2’s N protein, which is found inside the virus, wrapped around its RNA, the virus’s genetic code. It is thought the N protein plays a key role in helping the virus replicate after it has infected a cell. Portions of the N protein are very similar across all coronaviruses. And there are antibodies found within human cells that can recognize the N protein. Unlike the antibodies that latch on to the spike protein, these antibodies don’t prevent the cell from being infected. But when they find the N protein inside the cell, they break it into pieces, which the cell then displays on its surface. The body’s T-cells use these markers to identify the cell as infected and destroy it.

MyNeo is working with another Belgian company, eTheRNA, which has the ability to manufacture mRNA vaccines and also has created a proprietary adjuvant, which it calls TriMix that can significantly boost the body’s T-cell and B-cell response to a vaccine, Mike Mulqueen, eTheRNA’s head of business development, says. With help from MyNeo in selecting the right set of proteins, such as the N protein, to use with the TriMix in a vaccine, Mulqueen thinks they have a shot at producing a vaccine that will be much more effective against all future virus strains. The companies could have a vaccine in late-stage clinical trials within a year, he says.

Other scientists are skeptical

But there are skeptics of these approaches. Kathryn Hanley, a professor of biology at New Mexico State University who has worked on a dengue virus vaccine, says that there’s no reason to think that an inactive virus vaccine would create an immune response that is markedly different from what people experience with a natural COVID-19 infection. And in people with COVID-19, the spike protein seems to be principally responsible for the immune reaction.

This might be less of a problem for an mRNA-based approach, Mulqueen says. That’s because while it is true that in natural infections, the immune response to one particular epitope tends to dominate, if the body is presented with a different epitope in isolation—as could be the case with an mRNA vaccine—it is possible to elicit a strong immune response to that particular protein, particularly if an adjuvant is used.

John Moore, a professor of microbiology and immunology at Cornell University’s Weill Cornell Medical College in New York, isn’t so sure. He says that while the way the body forms an immune response to SARS-CoV-2 is still not very well understood, there are “so many bread crumbs that say neutralizing antibodies are what matters.” The T-cell and B-cell responses, which may be triggered by other proteins, could play a role, he says, but it is antibodies that are critical, and they form in response to the spike protein. “The S protein is the only target for neutralizing antibodies, so there is not a lot of choice,” he says.

Even if Moore is wrong, Mulqueen acknowledges one potential longer-term problem with MyNeo’s and eTheRNA’s approach to a universal COVID-19 vaccine: While in theory the well-conserved epitopes are more likely to be essential to the virus’s life cycle and thus less likely to mutate successfully, that might not be true. Those proteins have never been subjected to significant selective pressure. Once a vaccine is introduced that targets these other proteins, it is possible, Mulqueen says, that the virus will evolve a way to evade it too.

An age-old challenge

The history of medicine is littered with failed attempts to create universal vaccines. “People have been breaking their hearts trying to develop a universal influenza vaccine for a long time,” Hanley says. She says they have even tried the same idea—finding conserved epitopes across strains—and it has so far failed. Influenza viruses mutate far faster than coronaviruses, which have a step in their replication process that essentially proofreads the copied genetic code for errors, slowing the introduction of mutations. “Coronavirus is not influenza, so it may be possible, but it’s not a sure thing,” she says.

Some scientists also believe that any talk of the need for a new crop of vaccines to address the new variants of SARS-CoV-2 is premature. “So far a critical line hasn’t been crossed,” says Paul Offit, a pediatrician specializing in infectious diseases at the Children’s Hospital of Philadelphia and director of the Vaccine Education Center. Offit says crossing that line would mean people who were naturally infected with the original virus, or who had been immunized with the approved vaccines, were becoming reinfected with the new strains and becoming so ill from them that they needed hospitalization. He says he is encouraged by the high levels of antibodies seen in those who have received two doses of the current vaccines, particularly the mRNA ones from Pfizer and Moderna, and thinks those antibodies may be robust enough to continue to protect against severe disease even in the face of new strains.

He said he was hopeful, given that coronaviruses tend to mutate less than influenza, that if enough of the population can be vaccinated in the next several months—and given the fairly high number of people who have already had COVID-19—that transmission of the virus will start to drop off. With lower transmission, there is less chance for mutations that would be significant enough to warrant another generation of vaccines.

Moore shares this view, saying we’re not even sure that annual boosters to the current crop of vaccines will be necessary. “It is reasonable to plan and think about those things, but we’re not there yet," he says.

All the scientists, including those working on the next-wave vaccines, agreed that the key to preventing mutant strains even more troubling than the current U.K., South African, and Brazilian variants is to drive transmission of the virus down as low as possible by vaccinating as many people with the existing vaccines as quickly as possible. They say any discussion of not vaccinating countries’ whole populations and letting the virus run rampant through younger demographic groups, who are unlikely to become seriously ill from COVID-19, is a recipe for disaster.

“Resistant viruses arise under certain circumstances, and one of them is undervaccination,” Moore says. He also worries about extending the time between doses, as the U.K. has done (it is waiting 12 weeks between doses in order to give first doses to as many people as possible). That’s because there is no data from the Pfizer and Moderna vaccines on how long the immune response from the first dose lasts beyond the first four weeks. His concern is that if a virus finds a host with some immune response, but not enough to totally wipe it out and prevent replication, it puts selective pressure on the pathogen to find successful mutations.

Undervaccination is part of the reason why influenza has become endemic; even though flu kills more than 50,000 Americans each year, less than half of adults in the U.S. get the flu jab, Hanley notes. “Most public health programs become victims of their own success before they are truly successful,” she says. “That is why we have such trouble eradicating pathogens.”