Thread by @MAChristoLab

Sure, #posttranslational modifications rock our cells' world but where do they come from? We wondered for our favourite one and the answer floored us!

If you're also curious about the #evolution of #citrullination (and who isn't?🤓), here is the #preprint
disq.us/t/3pl33in

#Tweetorial to follow once #HomeSchool is finished for the day (your guess as to when that will be is as good as mine🎻🎻)

OK, here goes:

So when I started the lab, I never imagined we would carry out an evolutionary study. But as is often the case (and the fun!), Science takes you to fantastic new places. And this was quite out there!

We’d love to hear feedback/comments/criticisms on the work...1/

...or, if you’re interested but time-poor, here is a thread summarising our findings and how we got there. 2/

The citrullinating enzymes (PADIs, PADs) exist in all vertebrate genomes from fish onwards (fish have one PADI, humans have 5 paralogues) but are not in yeast, flies or worms. So we used to think of them as “vertebrate-specific”, which was always somewhat puzzling to me. 3/

What we initially set out to do was understand how the PADIs are regulated and part of that was looking for any sequence and structural features that might look like they can be subject to biochemical regulation. 4/

One discernible structural feature made us stop and think: it is an ancient fold and catalytic triad (found in bacterial agmatine deiminases) so what was it doing in these “modern” enzymes? 5/

We started digging a bit deeper and found that the catalytic domain of PADIs had homologous sequences across many bacteria and fungi. In fact, we found many PADI homologues in bacteria and fungi (but still not yeast, worms, or flies). 6/

The #phylogeny of the PADI gene was more unusual still and did not follow the known tree of life. Animal PADIs fall in a clade not with fungal (eukaryotic) sequences, but with homologues from two specific, relatively late-diverging #cyanobacterial clades. 7/

Fungal PADIs, on the other hand, are more distant and form a clade with other bacterial sequences, rather than animal ones. There is very strong bootstrap support and we get the same result regardless of how we built the tree (which method and with which portions of PADI) 8/

We also found that there are are two “types” of PADI sequence. A 3-domain one (like human PADIs), also found in cyanobacteria; and a 2-domain one, found in fungi and other bacteria. There are other synapomorphic features, some of which are functionally relevant 9/

And here was the most weird and wonderful finding of all (in my opinion): we looked at how much the animal-type PADI has diverged since cyanobacteria and we found that it changed much less than some of the most conserved proteins in life... 10/

...and even less than mitochondrial proteins, or proteins that are likely to have been subject to endosymbiotic gene transfer. 11/

Taking everything into account, we concluded that PADIs must have been transferred horizontally from cyanobacteria to animals. 12/

We produced the cyanobacterial protein and put it through various biochemical assays. It is a catalytically competent, calcium dependent deiminase, able (similar, or promiscuous enough) to citrullinate mammalian proteins (e.g. histones, which do not exist in bacteria). 13/

We therefore conclude that citrullination, as a catalytic function, arose in animals (probably at the base of deuterostomes) by horizontal gene transfer. Given how many aspects of human physiology and pathophysiology it impacts on, this was quite mind-blowing for us! 14/

Credit goes to Tom Cummings, formed PhD student in the lab for driving this (as well as doing some lovely biochemical work - watch this space!) and to our fantastic collaborators @kevin_gori, Luis Sancez-Pulido, David Moi, @cdessimoz and @CGATist 15/