We are in awe after watching Lucy Walker's animated documentary "The Contenders" about the Foldit group involved in mapping the structure of the Mason Pfizer Monkey Virus. Brilliantly done!
katfish 135 | Wed, 01/30/2013 - 18:28 | 4 comments )
Symmetric Design Results and Guidelines
It’s been a few months since we launched the symmetric design puzzles, and we now have a considerable database of results - your structures - to analyze. Among the designs are many creative, interesting, thought-provoking structural motifs. Unfortunately, the thoughts they typically provoke are: “Wow, this looks amazing! Too bad it probably won’t fold up correctly.” (To be fair, we’ve even made a few of them. No luck getting them to fold so far.)
This is partly our fault - we haven’t really given you a how-to guide; all you’ve had to go on are the energies the game shows you. And in minimizing the total energy, you’ve done well! But designing symmetric, multi-part proteins can be trickier than just looking at the raw energies, which is why we’re working on improving them, and making sure we inform you - the players - about what sorts of designs are most likely to work in real life.
With that in mind, this post will give some insight into what makes a well designed, realistic protein. We don’t want you to see this as a requirement to make ‘boring’ structures; we want to see more than just the same low-energy helical bundles or simple beta sheets that are already common in nature. The structures you’ve come up with so far have been awesome, and if you’re able to get an awesome looking structure that follows these principles, it’s much more likely it will fold up in practice - and we are much more likely to try it out in the lab!
Symmetric Protein Design Guidelines:
Symmetric copies are monomers too!
The first thing is to remember that, while symmetric, each copy of the protein has to fold up reasonably well by itself, as what we call a monomer.
This means that each non-symmetric copy of the protein needs to keep the principals of standard protein design in mind as well.
Each non-symmetric copy still needs a good hydrophobic core, and should try to minimize the surface area interacting with solution. Flat extended sheets are cool looking but have high surface area. If you look at the images below of an existing symmetric protein, you can see that each of the copies is a fairly reasonable looking protein, with a nice hydrophobic core that is highlighted in red in the image on the right. The purple region shows you the hydrophobic residues in the interface between the monomers, which allows these proteins to form a symmetric complex.
You can compare the above designs to something like the following symmetric protein:
This structure looks like it could have a chance to be stable as a complex, but individual copies with a sheet of strands is unlikely to be stable in solution. A lot of designs we saw looked stable as a whole complex but were unstable as an individual chain.
Why do the copies need to fold well by themselves? Well, living cells have quality control machinery to detect and destroy poorly folded proteins. If your monomer has a poor core it will fold more slowly, increasing the chance of being degraded before it can assemble in the complex you are trying to design.
Note that there’s an important exception to this rule: when the subunits are covalently attached to each other. Amino acid chains can be permanently glued together by covalent bonds, either by a disulfide (example: http://fold.it/portal/node/992625) or a small molecule that binds to the N terminus and acts as a “hub” (example: http://fold.it/portal/node/993779). In these cases, the chains don’t have to bump around in solution for a long time before finding each other; they’re stuck with each other for good, and the individual subunits don’t have to fold on their own. So you - the players - are free to interweave the chains any way you’d like; braided sheets, fully “communal” hydrophobic cores, etc. We’ll try to remind you which puzzles will involve these covalent hubs (where you don’t have to worry about independent chains being folded on their own) and which do not (where it’s important to make sure the individual symmetric pieces can fold at least part-way by themselves).
Long loops are too flexible to be controlled and they make your designs have higher chance to fold in unexpected ways. By shortening the loops, you can also make local interaction more likely to happen so that your protein can fold concurrently when it’s being synthesized. The reason for doing that is because free floating flexible loops or random coils get digested by other proteins in cells during protein synthesis. So, if you don’t want your designs to be eaten by other proteins, shorten the loops and make local interactions!
Continuing hydrogen bonds in the helix backbone gives you huge bonus of helix stability. A lot of time you may want to have a helix turn to another direction. In that case, we would recommend you to use a short loop to link two helices instead of building a broken helix with a kink.
Close sequence interactions
Having close sequences interacting with each other can make sure the protein can be partially folded and stabilized during synthesis. What this means is that is better to have beta strands pairing between close pieces of the protein. for example is better to pair strand 1 with 2 than 1 with 8.
Sidechains making hydrogen bonds with the loops.
Having loops forming hydrogen bonds with sidechains on stable strand or helix is a great way to stabilize your loops and the whole subunit so that your design is favored.
Hydrophobics in the interface, but not only hydrophobics.
There are different types of protein interactions but hydrophobic is the main driving force to bring proteins close to each other. However, having only hydrophobics at the interface is not a good strategy because except interaction, you also want specificity. Having only hydrophobics at the interface may allow multiple orientations of subunits that are coming together. Think of the hydrophobic patches as a lock and key - you only want to the correct lock patch to work with the corresponding key patch. A lot of people have experience in assembling foam mat, it gives a perfect example of interlock pattern that you may apply during interface design.
Hydrogen bonds between the subunits
Hydrogen bond can help stabilize interaction between subunits formed by hydrophobic interactions. It can also limit the number of orientation of subunits and increase interaction specificity.
We're about to introduce a new puzzle series with an awesome theme: clean energy. As many of you may know, hydrogen power is great. We can burn hydrogen to get energy, and the only byproduct is water. But where do we get the hydrogen from? Current technology does have the ability to extract hydrogen from various sources, but the processes for doing this aren't efficient enough for it to be practical. Our challenge is to use natural, renewable power sources to give us as much hydrogen as we want.
The Spiro Lab here at the University of Washington is designing a molecule which answers this challenge - a photo-electric hydrogenase catalyst. This molecule can absorb light to help power the ripping of hydrogen off of water. The molecule looks and acts a bit like heme, the thing that makes hemeglobin (and your blood) red. But instead of binding an oxygen atom, it will assemble and then release a molecule of hydrogen: H2.
So where does Foldit come in? Well, we need to deliver protons (charged hydrogen atoms) to the catalyst. To do this, we need to form a cavity around this heme-like catalyst that has space for a hydrogen atom, and lots of sidechains that can suck a second hydrogen atom out of the solvent like a sponge and bond it to the first hydrogen atom. Light will provide the power source, and a nearby electrode will allow electrons to flow to and from the catalyst. The protected pocket around the active catalytic site is a lot like what natural enzymes have, but in our case the central catalytic portion will be a molecule that we make ourselves.
For starters, we just want players to build all sorts of interesting structures around this symmetric molecule. Later on, we'll post puzzles that incentivize placement of hydrogen-bond-donating residues near the center of the catalyst, to help encourage hydrogen formation from water.
Try out the first new puzzle in this series now: http://fold.it/portal/node/993779( Posted by DrLemming 135 | Fri, 10/19/2012 - 22:52 | 1 comment )
Throwing a wrench at cancer
Wnt protein is widely recognized as a crucial component of vertebrate development. For more than three decades, scientists have sought to understand the structure of this important molecule. Unfortunately, obtaining a detailed understanding of Wnt's structure has proven to be quite difficult. After much work, a research group led by Dr. K. Christopher Garcia at Stanford University published the structure of Wnt bound to its target protein in June 2012. This binding event is partially facilitated by a fatty acid, an addendum to Wnt that is known to complicate molecular structure determination.
Wnt protein is secreted into the space around growing cells. It then binds to its target (Frizzled protein) on the surface of some of these cells. The attachment of Wnt to Frizzled leads to the transmission of a signal into the cell, which alters the development and physiological behavior of that cell by changing the way that the cell accesses the information in its DNA. This signaling event is modified by a number of supplementary proteins in the same pathway.
Our current project aims to replace naturally occurring Wnt with a surrogate protein through protein engineering methods. In a successfully engineered project, the newly created protein will show its ability to bind to the target (Frizzled in this case) during real life validation tests. We are submitting the structure of the target (Frizzled) and a potential binder Helix to you!
The creation of a successful Wnt-surrogate binder will signify an important advance in our ability to quickly employ emerging scientific data to facilitate the clinically-focused goals of our team of molecular engineers. Binding to Frizzled should allow us to interrupt the Wnt signaling pathway in a way that will be immediately useful as a tool for the many laboratories that are focused on human development. Looking farther into the future, control of the Wnt signaling pathway may allow us to limit the growth of Wnt-mediated tumors and may even prove useful in tissue engineering. Through the efforts of the scientists in the Baker Lab, our partners who dedicate their computing time to Rosetta@home, and Foldit players, we will begin testing the preliminary designs for a Wnt surrogate at our principal laboratory in Seattle soon.
Try out the new Frizzled Design Puzzle now:
Now that CASP10 is over, we've been able to analyze your Foldit predictions for the CASP10 targets whose natives have been released so far.
We won't know how well we all did overall until the CASP10 meeting in early December, which is when the assessors present their results. Until then, we can look at the natives that have come out already and at least see how Foldit compared to the Baker Lab's human CASP team (since we don't have the human submissions for any of the other groups that competed in CASP10).
Most of the targets whose natives have been released are from the refinement category, and we've noticed some very remarkable Foldit results!
Below are two plots for the same target, TR663, with all your Foldit predictions shown in green.
The black dotted line represents the starting model given to you in the puzzle (it's shown as a line because it scored so poorly, it would be off the plot).
In blue represents the native after wiggling/shaking with Rosetta (so closer to these blue points is where you want to be on both plots).
On the GDT plot you want to be as close to the right as you can (1.0 would be a perfect match to the native).
On the RMSD plot you want to be as close to the left as you can (0 would be a perfect match to the native).
In red are the 5 submissions by the Baker Human CASP10 Team, and in gold are the 3 submissions that we submitted using your Foldit predictions (the other two submissions were from the other TR663 Foldit puzzle).
You can see that many of your Foldit predictions improved the starting structure significantly, but we were very bad a picking these models out!
You'll also notice that our human Baker Lab CASP10 team was also unable to submit models that were significantly closer to the native than the initial refinement model.
You're probably wondering what the purple dot represents.
It is a Foldit prediction that was submitted by one of the three Foldit Groups that submitted their own predictions to CASP10!
Not only was the Foldit Group able to greatly improve the starting model, but they were also able to select that structure (out of all the top-scoring solutions from their Foldit Group) and submit it!
This is very exciting and it was not an isolated case for just TR633!
We always knew that Foldit players were amazing at folding, and now you are showing that you can excel at model selection as well (which, as you can see, is a very tough problem!).
We want to thank you all for your hard work during CASP10. We realize the protein size and puzzle load got a bit crazy at times, and we really appreciate your patience with it.
For all of you that just survived your first CASP, congratulations and welcome to the club!
For all of you who had been through it before (so you knew what you were getting into) and still put in so much time & effort this past summer: we can't thank you enough. Thanks for all the many hours of folding, as well as all the help you gave to the new players that had never gone through this before. We really appreciate all of you answering the many CASP-related questions and pointing players towards the relevant CASP-related websites.
Keep up the great folding everyone, we hope you all got a bit of rest after CASP10 and are ready for some exciting design puzzles that are coming soon!( Posted by beta_helix 135 | Fri, 09/21/2012 - 23:22 | 3 comments )