Even as governments and corporations push to declare an end to the COVID-19 pandemic, questions about the virus that causes COVID-19 remain unanswered. Will this virus, SARS-CoV-2, be tamed by the current array of vaccines? Or will these vaccines fail to keep pace with fast-evolving SARS-CoV-2 variants? And even if vaccines succeed initially, will they continue to protect us over the long term?
Questions such as these are unavoidable because SARS-CoV-2, like any other virus, will change ceaselessly to fulfill its sole purpose: survival. Essentially, SARS-CoV-2 has us caught between hope and worry—as dramatized by this article’s sidebar (see “The Temper of the COVID-19 Response, in Three Quotes”).
To help us deal with our uncertainty, we should look into the details of viral evolution, particularly with respect to SARS-CoV-2. We must not shrink from this task even if what we learn is discouraging. As the saying goes, “forewarned is forearmed.”
Drivers of viral evolution
“RNA viruses such as coronaviruses mutate at the rate of about one mutation every couple of weeks,” says Katherine Siddle, PhD, a postdoctoral fellow in the laboratory of Pardis Sabeti, PhD, at Harvard University. “Each time the SARS-CoV-2 virus infects, it has new opportunities to change. The number one thing driving the evolution of the virus now is the sheer number of cases globally,”
Most mutations are inconsequential. “Only mutations that benefit the virus allow it to become more prevalent in the population. But we must also consider immune pressure,” says Deborah Fuller, PhD, professor of microbiology at the University of Washington. Organisms generate immune responses that force a virus to vary its identifying features to avoid being recognized and eliminated.
“Transmission of a virus from an animal to human is another mechanism of viral evolution,” Fuller continues. “Once a virus jumps from an animal to a human, it is going to evolve to adapt itself to the new host.”
Coronaviruses, compared to HIV or influenza viruses, acquire mutations at a slower pace, partly because they have built-in correction machinery. Yet the highly transmissible U.K. variant acquired multiple mutations. Experts suggest that when a virus enters an immunosuppressed system, such as a cancer patient’s immune system, it lingers in the absence of adequate defenses and accumulates many mutations.
Viruses evolve not just through random and selection-guided mutations over generations (vertical evolution), but also through the mixture of genomes between members of different lineages (horizontal evolution)—the viral equivalent of sex. When multiple viruses infect the same host cell, hybrid progeny can be produced. This generally requires multiple species to be infected in large numbers, simultaneously.
“Fortunately, that’s not what we’re seeing with the SARS-CoV-2 variants,” says Paul Turner, PhD, professor of ecology and evolutionary biology, Yale University. “It is not so abundant in reservoir species that it is able to create this effect. We must monitor humans and animals around us for such events because we know coronaviruses can recombine.”
Animal experiments verify nonhuman primates, and some pets can be infected by SARS-CoV-2. “Minks can act as a reservoir of SARS-CoV-2, passing the virus between them, and pose a risk for virus spillover from mink to humans,” states the World Health Organization.
“In Denmark, [the risk for spillover] caused a panic, and the prime minister ordered all the mink farms destroyed,” recalls Laurie Garrett, a Pulitzer Prize–winning science journalist. “That was a costly intervention.”
Recombinant viruses also worry Esteban Domingo, PhD, a professor of research at Centro de Biología Molecular “Severo Ochoa.” “If, as it appears now, vaccines will prevent disease but not infection, recombinant genomic forms of SARS-CoV-2 are expected to become detectable epidemiologically,” he says. “Individuals infected by two different variants of SARS-CoV-2 may produce recombinant forms.”
Carriers who are not sick are estimated to account for more than half of all SARS-CoV-2 transmissions. Asymptomatic infection is linked to a mutation of a viral protein that compromises the virus’s ability to foil autophagy, an immune process that engulfs pathogens.
“If the virus is infecting individuals, and the individuals are asymptomatic, it is still expected that the virus is replicating within them,” Turner explains. “This means we are likely underestimating the virus’s evolutionary potential.”
Evolutionary patterns
SARS-CoV-2 variants indicate a pattern of convergent evolution. “We are seeing recurrence of certain mutations in different variants,” Siddle reports. “For instance, the mutation E484K has occurred multiple times independently, suggesting parts of the viral genome are susceptible to change or potentially advantageous to change.”
The U.K. and South African variants with multiple spike protein mutations show increased ease of transmission and resistance to neutralizing antibodies raised against the original strain. The African variant is also resistant to multiple monoclonal antibodies that target the receptor-binding motif.
“Everybody agrees that SARS-CoV-1 (2003) and SARS-CoV-2 are different strains,” Fuller points out. “The genetic and biological differences between the original SARS-CoV-2 that our vaccines are designed against and the new variants are almost at par with the differences between SARS-CoV-1 and SARS-CoV-2. It would not surprise me if down the line people regard these new variants as different strains.”
The receptor protein on human host cells that latches onto the viral spike protein is hACE2. This binding is so specific that species with a single difference in its residue are not infected by SARS-CoV-2. Some SARS-CoV-2 variants are evolving enhanced affinity for hACE2.
“This increases transmissibility because it allows the virus to ‘glom onto’ the cell more efficiently,” Fuller explains. “But with the B1.117, it also increases virulence in the sense that it causes greater mortality.”
Some patterns in the spread of SARS-CoV-2 cannot be explained through biology alone. COVID-19 has spread extensively in the African American communities in the United States, prompting speculation that some populations may be more vulnerable due to a combination of factors.
“The majority of people in sub-Saharan Africa are black or brown, and the virus has not done as well there,” Turner observes. “It is not color but the socioeconomic association of people of color together with comorbidities and lack of good healthcare that makes them vulnerable to the virus.”
Immune memory
Immunological memory keeps us safe from reinfection with the same pathogen, but it comes at a cost. Immune response upon exposure to a pathogen that is similar but not identical to the original pathogen can be less effective than a response prompted in the absence of memory. This “original antigenic sin” (OAS) results in age-specific severity of variant infection.
Studies show most humans have been exposed to other common seasonal human coronaviruses (hCoVs) before the emergence of the novel SARS-CoV-2 virus. However, this offers no benefit. The immune memory of hCoV antibodies does not protect against SARS-CoV-2 infections.
The mutation E484K in the B.1.526 lineage is now widespread in the northeastern United States, where COVID-19 first found a foothold in the nation. This indicates emerging variants may spread in regions where large sections of the population have antibodies to the original strain.
The current SARS-CoV-2 vaccines were designed based on the original strains. Meanwhile, the virus has been evolving. With the rapid emergence of several variants of concern (VOCs), the burning question is whether current vaccines will continue to be effective.
“You have to separate what the vaccines are designed to do from overall efficacy,” says Amesh Adalja, MD, FIDSA, senior scholar at Center for Health Security, Johns Hopkins. “The vaccines are designed to prevent three things: serious disease, hospitalization, and death. All three current vaccines (that is, the vaccines from Pfizer, Moderna, and Johnson & Johnson) do well at this even in the face of the variants.”
Estimates of vaccine efficacy depend on the objective. If the objective is to eliminate transmission and infection or reduce it to a point where infection is asymptomatic, the current vaccines will not be adequate.
“Antibody responses induced by the current vaccines are less effective in neutralizing emerging variants,” says Fuller. “Many manufacturers have already initiated production of updated versions of their vaccines that more closely match new variants.
“We’ve got these new variants, and our current vaccines are really not able to protect against infections or prevent people from getting a milder form of the disease. These people could still have viral replication in their body and continue to transmit it. If the virus is transmitting and replicating in the population, it could develop new mutations.
“Things may end up like flu. That is, we may have to update our vaccine annually to deal with genetic drift—the accumulation of minor modifications in the sequence of the virus. When drift occurs, immunity from the previous strain is less effective against the new strain. It is not completely ineffective, but it is less effective. That is why we update our flu vaccine every year—to keep pace with the virus’s evolution.”
Monoclonal antibodies, the basis of most antivirals, face the same issues as vaccines. “We’ve seen that even the [double]-antibody* cocktails, such as the one from Regeneron, can be less effective against new variants,” Fuller notes.
Universal vaccines offer a solution. “The idea,” Fuller explains, “is to generate immune responses against parts of the virus that are highly conserved across strains.” These critical pieces of the viral genome are central to its survival. Most vaccines, however, are generated against variable regions. This may appear paradoxical as it entails constant updates.
Fuller continues, “Viruses have evolved the ability to shroud important parts that they absolutely require from immune recognition. We could make a vaccine against these conserved sequences essential for the virus, but it is harder to generate an immune response against it.”
The trick to making a universal vaccine is zeroing in on a piece that is conserved plus immunogenic. “We’re trying to make them more immunogenic by modifying the sequence itself or using vaccinology technologies,” Fuller reports. “[These technologies include] new adjuvants or different vaccine modalities that make the virus more visible to the immune system. RNA vaccines incorporated in liponanoparticles and adeno-associated virus (AAV)-based vectors have in-built adjuvants because they are natural triggers of the innate immune system.”
Mapping the next pandemic
“It is urgent we solve the puzzle of where this damn thing came from,” Garrett declares. “Until we know the answer to that, we’re vulnerable. It is caught up in the political fight between the United States and China.”
Prominent genomic features of the virus also occur in other natural coronaviruses, making it unlikely that the virus is a laboratory construct. But we still need to figure out how a virus that existed in an obscure animal for millennia suddenly jumped across species to infect humans at a global scale. Once we understand how SARS-CoV-2 managed that jump, we will be better able to prevent similar jumps by other viruses in the future.
Quashing this pandemic and mapping the next will depend on rigorous global genomic surveillance of humans and animals. Currently, only a small section of COVID-19 patients is sequenced. Fast and accurate sequencing-based genomic surveillance helps public health officials track transmission routes, determine the rate of viral evolution, and understand if the virus is changing in ways that impact vaccine efficacy.
“We need to have more sequencing ability,” Adalja insists. “We must make genomic surveillance more routine. We have to continue to follow these variants in real time and understand how they change their function.”
“Increased accessibility to technological advancements has paved the way to curb the spread of viruses,” says Arvind Kothandaraman, general manager of specialty diagnostics at PerkinElmer. “Large datasets from different sources present several challenges—unification, harmonization, and accurate interpretation. Next-generation sequencing has enabled rapid biological pattern recognition due to the unprecedented rate at which genetic information can be shared. This kind of collaboration is an unlikely but welcome byproduct of the pandemic.”
“The problem,” Domingo complains, “is that it is unlikely that a rate of transmission will be calculated soon enough with results that will convince health authorities to implement additional restrictions to the human population. The data should be dramatic to reinforce a stricter confinement in many countries at the same time.”
Basic mechanisms of viral evolution, which remain unclear, require careful study. “Using any biosafety level 3 virus is not practical or safe,” Turner notes. “I am optimistic we can find suitable models to gain some ground.” Kothandaraman adds, “We cannot take our foot off the gas now. We must remain agile, pivot as needed, and continue to invest in our R&D engine to help outpace the virus.”
Fortunately, there are powerful technologies that are being widely deployed, and there is an unprecedented willingness to collaborate among scientists in academia and industry. These trends will help us get ahead of SARS-CoV-2 and combat evolving pathogens.
*The text has been modified to clarify that Regeneron’s REGEN-COV is a double-antibody cocktail. In a recent update, the FDA has affirmed that REGEN-COV maintains potency against all the major variants (UK, Brazil, CA, NY, SA).