Read on Twitter
@BenButt2, Edward (Twitter-free) and I have just published a protein structure that is definitely wrong. Curious result, but a fun story, so here goes…
https://wellcomeopenresearch.org/articles/5-154
https://wellcomeopenresearch.org/articles/5-154
Many moons ago Edward and I were investigating the C-terminal domains of proteins from the HOPS #membranetrafficking complex 
. We purified the C-terminal domain of VPS39 and attempted crystallisation, but nothing happened and the lab moved on…
. We purified the C-terminal domain of VPS39 and attempted crystallisation, but nothing happened and the lab moved on…
Fast forward 2.5 years and I’m clearing out the crystallisation incubator. I inspect all the plates before chucking them out (because I’m like that) and I see this big beautiful crystal!
We head to @DiamondLightSou I04 where a fluorescence scan shows signal for Zn (not surprising, this domain is predicted to be a Zn finger). We collect 2wl MAD data, I solve the structure and an initial trace reveals a weird beta-barrel like fold. Definitely not what I expected 
Closer inspection shows that the Zn atoms are bound by His side chains *from the purification tag*. D’oh!
Looks like the protein mis/refolded during the super-long crystallisation experiment.
Looks like the protein mis/refolded during the super-long crystallisation experiment.
I meant to write this up as a paper – it was my summer project for three summers running but I never got round to it …until #COVID19 happened. Cue the lab is all stuck at home and @BenButt2 was looking for something fun to work on.
What better than a new structure to while away the hours at home! Ben solved the structure (no cheating – I didn’t let on about the weirdness!) and refined it, and we wrote up the paper all via the internets. It’s published in @WellcomeOpenRes, currently awaiting open peer review
But the coordinates and maps are available at the @PDBeurope ( http://doi.org/10.2210/pdb6ZE9/pdb) if you want to take a peek.
All the original images are online too: https://doi.org/10.17863/CAM.53867
All the original images are online too: https://doi.org/10.17863/CAM.53867
TL;DR This isn’t what the VPS39 C-terminal domain looks like, but hopefully someone will find it useful
It's perhaps best to avoid His tags for Zn fingers (or chop off the tag before crystallisation). Also, remain sceptical if you solve a structure from 30 month old crystals!
It's perhaps best to avoid His tags for Zn fingers (or chop off the tag before crystallisation). Also, remain sceptical if you solve a structure from 30 month old crystals!
