Thread by @fried_lab

Why is there so much disorder in eukaryotic proteins? What is its relationship with condensates? Why can some proteins be saved but others get degraded? After years in the making, we think we have answers and they're all connected. Excited to share a 🧵:
biorxiv.org/content/10.110…

First some background. 2 y ago we developed a way to measure which proteins can put themselves together after being unfolded with denaturant using LiP-MS. tl;dr some can, some can’t (b/c they need chaperones or prefer to fold on the ribosome). 2/10.
pubs.acs.org/doi/10.1021/ja…

We unfolded yeast's proteome and challenged it to refold spontaneously. The result? 🥁... Yeast proteins are surprisingly efficient refolders, much BETTER than E. coli's, despite being larger, having more domains, & more β content. How can this be? 3/10

The secret is disorder. The more disorder a protein has the better it refolds (like a built-in chaperone). Yeast proteins have higher levels of disorder than E. coli's, and this neatly explains why they refold better. Same goes for N. crassa, another fungus. 4/10

But why, you ask? Our data suggest a simple model. Look at all 2-domain proteins. If domains are right next to each other➡️poor refolding. If domains are separated with a long linker➡️they refold as well as single-domain. So disorder evolved to help proteins refold. But why?

Disorder is frequently connected with LLPS. Sure enough, if we look at stress granules, proteins that go into SGs during heat shock (beautifully assessed by @DrummondLab, @ewjwallace) are also generally excellent refolders. (Same goes for P-bodies). 6/10

This makes sense b/c upon dispersal SG proteins are NOT degraded, but instead are returned to cytoplasm through Hsp104. Hsp104 can pull proteins out of SGs, but unfolds them in the process. Retrieval is efficient though since Hsp104's natural clients are refoldable. 7/10

Some condensates contain disordered proteins but others don't. Why? Disorder makes proteins more refoldable, which makes them easier to extract from condensates. So dynamic droplets need disorder, but longer-term droplets that get cleared by degradation do not. 8/10

When we started, we knew very little about IDRs, condensates, Hsp104. It has only been possible thanks to amazing friends/mentors (@dallandrummond, @alexholehouse), and the trends in our datasets which kept telling us there HAD to be some reason for all of this! 9/10

There's loads more in the preprint; check it out and please give us your feedback! This project was >2 y of hard work 💪 by @PtoChem & Atharva Bhagwat w/ contribs from other students the @fried_lab. Thanks to @NSF 🧪@NIH_CommonFund ⚕️@HFSP 🌎for support.