The optimistic mood around COVID-19 vaccine rollouts has been clouded by new variants of the virus, which could trample the efficacy of vaccines or escape them entirely. Already, recent trial results from Johnson & Johnson and from Novavax suggest that a variant that first arose in South Africa (B.1.351) and probably a variant identified in Brazil (P.1) are partially escaping protection provided by their vaccines. 

Specifically, mutations in the viruses’ spike proteins allow them to avoid being bound by antibodies produced after vaccination or natural infection. “The South African variant appears to partially escape antibody responses,” says Dan Barouch, the director of the Center for Virology and Vaccine Research at Beth Israel Deaconess Medical Center in Boston. With novel variants expected to emerge, the question is, which vaccines would be quickest to rejig and manufacture if updates become necessary?  

A few vaccine makers have already announced that they are gearing up for second generation vaccines. “It may not be necessary, but is probably good insurance to buy right now,” says immunologist Alessandro Sette of the La Jolla Institute for Immunology. 

A quick fix

Of the common approaches to design—or redesign—a vaccine, mRNA vaccines are the most expeditious. They pack the genetic recipe for the SARS-CoV-2 spike protein inside a nanoparticle, which is engulfed by cells that then manufacture the protein and tutor the immune system to recognize it. “They are using chemically synthesized mRNA. There’s no cells needed, so it is so much quicker to make and then purify,” says immunologist Sarah Caddy of the University of Cambridge. 

Moderna designed its mRNA construct last January and had it ready for testing by March 2020. Pfizer/BioNTech’s mRNA vaccine was not far behind, and was the first to gain emergency use authorization from the US Food and Drug Administration, beating Moderna by one week last December. 

It wouldn’t surprise me if every vaccine maker is now making versions that will tackle the Brazilian and South African strain, and [the] UK [strain] for good measure.

—Luke O’Neill, Trinity College Dublin

“The simpler the vaccine, the easier and faster it would be to ‘update’ it,” Wolfgang Leitner, head of the innate immunity section of the National Institute of Allergy and Infectious Diseases (NIAID), tells The Scientist in an email. “Changing an RNA-sequence would ‘only’ require switching out the plasmid that’s used as a template to make the RNA, and modifying a sequence in a plasmid is also quite simple.” Indeed, Moderna was quick off the blocks again, announcing late last month that it had a candidate booster vaccine encoding B.1.351’s spike protein ready to be tested in a clinical trial in the US. 

Arguably the next quickest vaccine type to manufacture is the vector-based variety, which delivers the spike protein recipe using a live virus. To make the University of Oxford and AstraZeneca’s product, for example, living “producer cells” from a human cell line are engineered to become mini vaccine factories, before being burst open to release a weakened adenovirus vector carrying the spike protein gene. “That will take about twice as long [as mRNA vaccine production], so three months to make an AstraZeneca-type vaccine,” says Luke O’Neill, an immunologist at Trinity College Dublin. But according to Caddy, updating vector-based vaccines should be as straightforward as revamping mRNA ones. “The sequence just needs to be inserted into the adenovirus sequence, then the adenovirus grown and purified.”

Another option to fight off mutants is to change not just the spike sequence delivered by vector-based and mRNA vaccines, but to add additional proteins from the virus to their cargo. Caddy advocates for including the nucleoprotein that wraps around the virus genome in COVID-19 vaccines. “The nucleoprotein doesn’t mutate anywhere near as much,” she says, “and mutations there are more likely to make the virus nonfunctional.” 

“There are a couple vaccines in the pipeline (but far from being ready for prime-time) that encode additional viral proteins,” notes Leitner. “The challenge is that they are more complex and, therefore, take longer to develop.”

Barouch agrees. Adding nonspike proteins would “increase complexity and slow down timelines,” he says, so will likely not be done unless the spike-only vaccines fail.  

All about that spike?

While adding additional antigens to mRNA and vector-based vaccines would slow down production times for these rapidly produced vaccines, another popular vaccine approach already has the advantage of triggering the immune system to recognize multiple parts of the virus: inactivated SARS-CoV-2 viruses. This approach involves producing large amounts of virus in cell culture, then treating the viruses with a chemical to render them no longer infectious. This process can take more time—not to mention the challenges of standardizing inactivation and conducting essential quality checks to ensure the virus is indeed safe—and so vaccines of this type have lagged behind mRNA and vector-based vaccines. But the upside is that the immune system sees an entire virus, not just the spike protein. 

“Even if there are spike mutants, there will be other antibodies that [inactivated vaccines] will bring onto other proteins in the virus,” says O’Neill. For that reason, these vaccines are “more likely to be valuable in the face of variants,” adds Caddy, “because you get a response against other proteins, which are not mutating as quickly as the spike.” 

If and when these upgraded vaccines will be needed is still a big question. 

Leitner adds that producing new vaccines for COVID-19 would simply involve growing a new variant. “When it comes to inactivated viruses, I would actually rank them among the very easy ones to ‘update,’” he says—“as long as the variant grows just as well (maybe even better) as the original virus.”  

Not everyone agrees that targeting more antigens will be helpful. Immunologist Danny Altmann of Imperial College London says T cell and B cell immunity is elicited from many parts of the wildtype virus, “but I can see no hard evidence that there’s strong correlates of protection outside of spike neutralization.” In other words, vaccines that target the spike protein are the most likely to offer protection against SARS-CoV-2 infection. 

Immunologist Kingston Mills of Trinity College Dublin agrees that neutralizing the spike is crucial, and that other targets are of peripheral importance to COVID-19 vaccines. “Nonspike proteins may generate T cell responses that may help, but won’t generate neutralizing antibodies, and antibodies against N [nucleoprotein] or other proteins will not have a role in protective immunity,” he says.  

For now, then, vaccine makers are focused on updating the spike protein sequences in their shots to match the variants that appear to be partially evading SARS-CoV-2 antibodies. Even for protein-based vaccines—the fourth major strategy for making vaccines for COVID-19, involving the production of the spike protein in the lab, which in Leitner considers to “be the most involved process of them all”—companies are already moving on a second generation of shots. In January, for example, Novavax announced that it started developing new versions of its protein-based vaccine against emerging strains. “Animal studies are underway and we could be in the clinic as soon as a month or two from now,” a spokesperson from Novavax informs The Scientist in an email.

“It wouldn’t surprise me if every vaccine maker is now making versions that will tackle the Brazilian and South African strain, and [the] UK [strain] for good measure,” says O’Neill. 

How soon are vaccine updates needed?

If and when these upgraded vaccines will be needed is still a big question. There are two main strands of evidence to consider in determining how newly emerged variants are affecting vaccine efficacy. First, the most recent clinical trial results: Novavax achieved more than 90 percent effective in preventing COVID-19 in a Phase 3 trial in the UK, according to an interim analysis, but just 60 percent effective in a smaller trial in South Africa where B.1.351 is rife. For its adenovirus vector–based vaccine, Johnson & Johnson reported efficacy of 72 percent in the US but 66 percent in Latin America and 57 percent in South Africa, where troubling variants are more prevalent. The conclusion is that these spike mutants indeed seem to be reducing the protection that vaccines can provide.  

Second, the lab studies: To explore the effects of variants on the Moderna and Pfizer/BioNTech vaccines, researchers took serum from vaccinated people and tested how it fared against the new spike variants. These experiments showed that antibody responses declined six- to eightfold. But because of the high antibody titers the vaccines generate, this is still in the protective range, according to Barouch, who was involved in the development of Johnson & Johnson vaccine. These in vitro results suggest that these vaccines elicit high enough antibody levels to neutralize the new variants.

However, the challenge posed by new variants was highlighted by an announcement that similar lab studies found the AstraZeneca vaccine to provide minimal protection against mild and moderate COVID-19 upon infection with the B.1.351 variant, far less than the protection it provides against the original SARS-CoV-2 strain. The Financial Times reported that it had seen trial results to support this finding, though the vaccine’s effectiveness against severe disease has not yet been determined. South Africa paused rollout of the vaccine following the announcement. Sarah Gilbert at the University of Oxford says in a press release that efforts to develop a new generation of the vaccine are now underway.

It will be difficult to know from serum samples if or when a mutant has escaped a vaccine, partly because researchers do not know where to draw the line between an immune response that is protective against SARS-CoV-2 and one that is not. “We don’t know for sure exactly what kinds of immune responses or what antibody titers are truly necessary for clinical vaccine protection in humans,” says Barouch. When he and his colleagues transferred convalescent serum to macaques and challenged them with SARS-CoV-2, they found that relatively low antibody levels protected the primates, but only if the serum contained adequate levels of CD8+ killer T cells.

The involvement of these T cells, which kill cells infected with virus, complicates the picture. Having more CD8+ T cells is linked to milder COVID-19. So far, from his calculations, Sette says the T cells stimulated by vaccines or infection to an early lineage of SARS-CoV-2 should largely recognize the new variants. He studied T cell responses in COVID-19 and found that an average person’s T cells recognize multiple parts of SARS-CoV-2—“at least 15 to 20 different pieces,” he says, including on the spike and several other proteins.  For that reason, he concludes, “it is very unlikely that the virus could mutate to escape T cell recognition.”

Some vaccines are suspected to provoke a stronger T cell response than others do. Vector-based and mRNA vaccines, for example, mirror natural infection with SARS-CoV-2, forming the spike protein inside cells—a process that is thought to strongly activate killer T cells, says Barouch, whereas “protein-based and inactivated-virus vaccines generally don’t raise a strong CD8 T cell response.” Chemical adjuvants used in protein and inactivated-virus vaccines may promote killer T cells, says Sette, and side-by-side comparisons of T cell and antibodies have not been done yet between vaccines.

“There could be differences among [vaccine] platforms in terms of different immune responses, or different levels of T cell responses,” says Barouch. “That could be why some platforms may be better than others for tackling variants. But that is theoretical, and we don’t know that for sure.”