ACE2 variation may affect spike protein binding

14 May 2020

Interview with 

William Gibson, University of British Columbia


Human genetics seem to affect COVID-19 infections; but what about the gene for the crucial protein, ACE2? Does variation in the ACE2 gene make the coronavirus see you as a bullseye - or even the opposite, might it protect you? Medical geneticist William Gibson at the University of British Columbia tried to find out, using a big database of people’s DNA called the gnomAD, and told the story to Phil Sansom...

William - There's just over 141,000 people represented in the gnomAD database.

Phil - And are you looking at this ACE2 gene? What it's like for all of those people?

William - We did. Not every person had reliable data at every part of the gene. Fortunately, the gnomAD database gives us data on the quality of each person's DNA sequencing at each specific site, so we were then able to look at the number of X chromosomes in both males and females that had reliable data at each of the sites that we were looking at.

Phil - Oh, okay. So you've got all your people, you've got twice the number because all the women have two X chromosomes, and then you've whittled down to get all the ones that are definitely accurate. What differences did you find in this sort of final amount of data?

William - In the final analysis, we found that there are differences in the ACE2 protein and these differences seem to be distributed differently among different human populations and predicted to be at different levels between males and females.

Phil - Talk me through this. What kind of variants can you get?

William - You can get genetic variants that are predicted to stop the protein from being made. They either break the protein or they make a protein that's too small to do anything. You can also get variants that are predicted to change the shape of the protein but not break it so the protein likely works pretty well and we call those missense variants. We're looking at a small number of the variants that are specifically predicted to bind to the virus as opposed to all of the missense variants overall.

Phil - How many of these types of variants did you actually end up finding?

William - Between 14 and 15 missense variants in the ACE2 protein that were predicted to bind the virus. Some of them were predicted to bind the virus more tightly and some of them were predicted to bind to the virus less tightly.

Phil - How common are these differences overall?

William - Uncommon but not unheard of. Among males of European descent, around 1 in 170 males would be predicted to have one of these variants. It's about 1 in 80 females. Interestingly enough, these variants were more frequent in individuals of Ashkenazi Jewish descent, and they seem to be less frequent among Latin Americans, South Asians, people of African descent, people of Finnish descent, people of East Asian descent.

Phil - Really what does that mean?
William - It's not entirely clear what it means. What we're looking into now is whether these variants actually affect susceptibility to disease and disease outcome. We're not sure about that yet.

Phil - Did you find that, for example, the Ashkenazi Jewish population had more of the variants that bound the virus better or more of the variants that bound the virus worse or neither?

William - I confess, we haven't actually split out the numbers that way. Partly because we're not really sure how good our prediction software is. In other words, do the predictions that we make using the software actually correlate to actual binding of the protein or not?

Phil - If they did, is there anything you could do about it?

William - Theoretically, if someone were found to have a high susceptibility variant, then that be very useful for them to know in terms of either self isolating or of course for their healthcare practitioners if they were found to then contract the virus. That's someone who would need quite carefully managed care.


Add a comment