You are here

Beta-Hairpin Design

3 posts / 0 new
Last post
Beta-Hairpin Design


I have performed a wet-lab study about designing beta-turns with two different sequences and found that both sequences provide a huge difference in the folding rate. Is it possible to model the same with Rosetta, I mean folding is not possible but based on energy can this be studied ??

Simultaneous design and folding of a hairpin is possible with Rosetta or not ??


Post Situation: 
Wed, 2014-02-12 23:18

One thing you'll want to be careful with is the thermodynamics/kinetics distinction. Rosetta is mostly set up to model stability. What kinetics it does model (e.g. enzyme reactions) it does by recasting it into a stability question (e.g. transition state stabilization).

So Rosetta probably wouldn't necessarily be set up to model differences in a purely kinetic folding rate difference. To the extent it would do so, the arguments would likely involve the stability of putative folding intermediates (e.g. are the energies of the intermediates along the putative folding pathway better tuned with each other in the better mutant) or perhaps the stability of dead-end side products (e.g. does the slower folding variant have kinetic traps that it gets stuck in, slowing it down.) You can certainly pursue an investigation of folding rates with Rosetta, but you'd likely have to justify why you didn't just use molecular dynamics, which has a better conceptual match to kinetics-based arguments than does Rosetta.

If you're modeling just an isolated hairpin, the other issue you'll run into is that Rosetta is set up to model larger globular proteins. If you have just a hairpin, the Rosetta energy function might not be as predictive as you would like, as there isn't a hydrophobic core. (This argument disappears somewhat if the hairpin is part of a large protein.) Also, Rosetta has traditionally had a much better time modeling alpha helices than beta sheets. It certainly can be done, but in small systems like a beta hairpin where that single beta-beta interaction is critical for structure, you need to be careful that the protocol you're using gives you reasonable results.

Design is a different story, and Rosetta does design easier than something like a molecular dynamics based system. The key here is to again recast your design arguments in the light of stability. (I want to increase the stability of *this* intermediate, while decreasing the stability of *this* off-pathway structure, all while making sure that *this* intermediate stays at a lower energy than *that* one.) You could do things like put together your various intermediate structures and use multistate design ( to optimize the sequence to fit all the competing criteria. The key to success is how well the stabilities of your putative intermediates match up with the actual kinetics of folding.

Thu, 2014-02-13 08:36

Thanks for such a useful insight into the topic. Well, after performing equilibrium stability studies, I found that the mutation doesn't affect the stability much. On the other hand, the folding rates are drastically affected. I am really interested to find out how a change in merely two residues of the protein resulted in such a large difference in the folding kinetics. I suspect it's related to the way the protein folds. I trying it with REMD also, but wanted to know whether Rosetta could of some help or not. Actually, we performed the experiment by first designing the sequences using Rosetta and selected the ones with lower Rosetta energy score. We found that it's difficult to make a correlation between Rosetta energy and folding kinetics.


Thu, 2014-02-13 16:00