Almost from the moment it made the jump to humans, the SARS-CoV-2 virus has been picking up mutations and creating new lineages as it expands into different populations. In practical terms, the vast majority of these mutations makes absolutely no difference; the resulting virus has the same properties as the unmutated form it’s derived from.
But there have been a number of cases where variants surge in frequency. Early on in the pandemic, this was often the product of the variant moving into a previously unexposed population—a matter of chance rather than a feature of the virus. Separating out these cases from instances where mutations make the virus more dangerous is a serious challenge. But this week, an international team of researchers has published evidence showing that a variant first characterized in Brazil is likely to represent a significant additional threat.
There’s a lot of uncertainty about the details, but the virus appears to be more infectious and more likely to infect those who have immunity to other viral strains, and it might even be more lethal. And, as of when the paper was written, the lineage had been detected in over 35 countries.
The second wave
Earlier this year, we described the situation in the Brazilian city of Manaus, which the first wave of coronavirus infections had hit hard. But that was followed by a long period of low infections, despite an indifferent response to the pandemic by the Brazilian government, leading some to suggest that the city might have reached a level of infection sufficient to provide herd immunity.
That hopeful thought was brought to an end in December, when a second wave of infections started up in the city, straining its health care systems and causing another surge in deaths. The infection rates were so high that it raised the suspicion that there might be a new strain of virus that could evade the immune response generated by infections that occurred during the first wave.
Brazilian healthcare workers responded to the rise in cases by sequencing the genomes of some of the viruses causing the second wave of infections. Prior to this second wave, only seven viral genomes had been obtained from Amazonas, the state where Manaus is located. The new effort increased that number by 184, although not all of these were complete genomes.
The genomes revealed the presence of a lineage researchers call P.1, which is an offshoot of a strain that had been present during the first wave. Since then, P.1 had picked up a large number of mutations, including 17 individual mutations that altered the amino acid sequences of the proteins it encodes, one insertion of new bases, and three deletions of bases. That’s a substantial number of changes and suggests a high level of mutations picked up since March. Timing estimates suggest that P.1 originated in November, just before the start of the large second wave in Manaus.
Over the course of the second wave, the P.1 variant went from not being detectable in the samples taken to accounting for 87 percent of viruses sampled just seven weeks later. Viral genomes from elsewhere in Brazil indicated it was also spreading rapidly within the country, showing up in cities that were on popular flight routes originating in Manaus. This indicates that P.1 likely originated in the city.
What is this thing?
Tests for the virus that use polymerase chain reaction (PCR) involve a cyclical amplification of the virus’s genome. As a result, if you start with more viral genomes, you’ll reach a detectable level of signal in fewer cycles. This is thought to mean that the cycle count needed for detecting the virus provides a rough measure of the viral load carried by the person the sample came from. In the case of the P.1 strain, tests showed a fairly consistent, if small, indication of increased viral load.
Because the samples came at different times after infection, however, the researchers can’t tell whether this is indicative of higher maximum levels of the virus or a longer infection duration. Neither is especially good.
To try to understand how P.1 might have influenced the second wave of infections in Manaus, the researchers developed an epidemiological model that allowed them to track two different strains of the virus. The first strain was set up with the typical properties of SARS-CoV-2. For the second, they were able to adjust the properties of the virus, such as the immunity provided by prior infections and its transmissibility. This let them determine which properties were consistent with the dynamics of the second wave in Manaus.
Overall, the model suggests that P.1 is very likely to be more transmissible than prior strains of SARS-CoV-2, and it’s likely to be roughly about twice as infectious. There’s also an indication that it can evade the immune response generated by past infections to some extent. The model suggests there’s at least a 10 percent chance that the variant can evade immunity, but it’s unlikely to be more than a 50 percent chance.
There was some evidence of enhanced lethality due to infection by the P.1 strain. But the timing of the strain’s rise was such that the evidence came from a period where the hospitals were on the verge of being overwhelmed. So the authors are treating this possibility cautiously.
What might be causing these changes? At least 10 of the mutations seen in the P1 strain affect the virus’s spike protein, which the virus uses to latch on to cells it infects. At least eight of those mutations seem to have been selected for over the course of the strain’s evolution, suggesting they assist in making it more infectious. Three of the specific changes have also been seen in another lineage of virus that has caused concerns, and at least one of them has been shown to interfere with antibodies that attack the virus.
So while this data isn’t really a decisive indication that P.1 poses a distinct threat to us, it’s all certainly consistent with that concern. And it would help explain why Manaus had two distinct waves of infection that seem to have hit a substantial fraction of the city’s population. Still, as the authors of the new paper point out, we don’t fully understand the consequences of mutations that alter proteins targeted by antibodies. Until we get a grip on that, we won’t really know how worried we need to be about P.1 and other variants.