Thread by @sigallab

Our work on the divergence of beta and delta variants into serological phenotypes has just been posted to medRxiv:

medrxiv.org/content/10.110…

Viruses such as dengue and polio have "serotypes". That is, they have circulating strains which have diverged so much from each other in terms of the antibodies they elicit that vaccination with one strain may not protect you from the others. Will this happen with SARS-CoV-2?

We mapped how the antibody response to one variant deals with a different variant. To do this, we used ancestral (originally circulating), beta and delta viruses,

as well as antibodies in the blood of people infected with SARS-CoV-2 by the ancestral viruses and either beta or delta variants in South Africa during one of the country’s three infection waves.

Therefore, antibodies from people infected by the ancestral circulating viruses were used to neutralize (inhibit) the beta and delta variant viruses, antibodies from beta variant infected people were used to neutralize ancestral and delta virus,

and antibodies from the delta variant infections were used to neutralize the ancestral and beta virus.

We also mapped a virus which evolved from an ancestral virus in a person with advanced HIV, who could not clear the SARS-CoV-2 infection for six months. We call this evolved virus D190, since it was from day 190 post diagnosis.

This is the map we got

As you can see, the beta and delta variants are far removed from each other in terms of the antibody response: delta escapes from beta infection elicited antibody immunity (12-fold lower neutralization) and beta dramatically escapes from delta elicited immunity, a 34-fold drop!

The ancestral virus was at the hub. The beta escape from delta immunity was greater than anything we measured before. We asked Penny Moore to confirm it using a different assay, and it was confirmed.

The D190 virus which evolved because of immune suppression behaved very much like beta, despite multiple genetic differences in the spike region.

It is tempting to design vaccines specifically to the delta variant, since it is so dominant now, or maybe another future variant. But, these results may suggest that if you do that, you may open the door to the other variants, beta or beta-like, in the case of delta.

One option is to go for the hub - ancestral, since it is serologically closer to the diverging variants than they are to each other. Or, include multiple serologically divergent variants in the vaccine.

I want to acknowledge @Sandile_Cele22, the driving force behind the experiments, @farinakarim and @KhadijaKhan24 who make the cohort and lab work possible, our terrific clinical partners Yunus Moosa and Bernadett Gosnell, and @rjlessells, @houzhou, and of course @Tuliodna