Puzzle 539 : Beginner Immunotherapy Design Puzzle
Foldit Player Mark-'s solution for Puzzle 539 A pretty creative solution where the Trp resides in the cavity and the Leu and Ile make nice hydrophobic contacts (and two backbone hydrogen bonds) with the three prolines in the loop of CTLA-4. This is especially relevant since these prolines make up a part of the native binding site.
Foldit Player JackWeaver's solution for Puzzle 539 Instead of a Trp, a Met is placed inside the pocket which contributes to a favorable binding energy of the whole complex. The Trp is stacked against the hydrophobic proline loop of CTLA-4, which again makes it an interesting solution.
Foldit Player Flagg65a's solution for Puzzle 539 This solution is also intriguing since it combined the use of a Leu interacting with the prolines from JackWeaver's solution and the Met inside the cavity from Mark-'s solution. Furthermore a nice stabilizing hydrogen bond is formed between the backbones of the Met & Tyr residues and a proline in the CTLA-4 loop.
[23:00:53] <beta_helix> Hi everyone, we'll be starting the Scientist Chat in a few minutes (just setting up the new puzzle so you can all check it out and ask questions). After the Scientist Chat (~30 minutes) jflat06, Seth and I will talk about some other Foldit stuff and we can answer questions then!
[23:01:32] <frood66> thx beta
[23:01:47] <@CFC> hey beta, hello evdh :)
[23:05:34] <@beta_helix> Ok everyone, let's get started! I'd like to welcome Eric who has just joined the lab and is working on this very exciting project! It should be in Gallery Mode now for you to check out!
[23:05:39] <@EvdH> Hi guys!
[23:05:52] <@beta_helix> (The puzzle will go up once 535 closes)
[23:05:52] <frood66> hi eric :)
[23:06:02] thomirc8 too late!
[23:06:25] <@CFC> :) we are almost incandescent with enthusiasm!
[23:06:36] <frood66> haha cfc
[23:07:54] <@beta_helix> Eric will start by answering Susume's question in the blog:
[23:07:54] <@beta_helix> http://fold.it/portal/node/992285#comment-16647
[23:07:56] <urlbot> Long URL from beta_helix has been shortened to http://tinyurl.com/co3xxjx
[23:08:20] <@EvdH> @susume I've just read your question at the blog post. The system indeed seems a little bit complicated from the outside. But the CTLA-4 protein is expressed by T-cells. And to kill a tumor we want the T-cells to be as active as possible, one way of doing that is by making sure the CTLA-4 pathway cant be activated.
[23:09:50] <Susume2> ah, thank you
[23:11:34] <frood66> so we are blocking? - please pardon my ignorance
[23:12:21] <@EvdH> yep! If you guys are able to bind something to the CTLA-4 protein we can block that pathway :)
[23:12:45] <frood66> ok - nice to know :)
[23:13:00] <@EvdH> Since we posted it as a gallery puzzle you are now able to view the structure
[23:13:12] <TimovdL> How do you make it selective to only do that in tumor cells?
[23:13:49] <@beta_helix> So please let us know if you have any questions about the gallery puzzle, or what you need to work on.
[23:14:48] <@beta_helix> there should be if you pull it too far, we made them generous to start with!
[23:14:49] <Susume2> is it possible to show bondable atoms on the frozen target protein?
[23:16:09] <@EvdH> Good question Timo! That's a problem currently dealt with, because there are two antibodies available which are able to bind CTLA-4 already and they are used in clinical trails. But as you already pointed out they suffer from a lot of side effects since they are not only active near the tumor but throughout the whole body. So a lot of side effects will come up. If we can use a much smaller protein to block the CTLA-4 we can fuse it to a tum
[23:16:10] <@EvdH> specific tag for example.
[23:16:41] <@CFC> since CTLA4 is only a part of the regulatory system in CD4 activity, is there a 'puzzle strategy' for this nest of puzzles?
[23:17:08] <@beta_helix> @Susume2 it doesn't look like that is possible (was that also the case for the Flu Puzzle?). Hopefully that is not to hard to fix, as that would obviously be useful!
[23:17:20] <Susume2> it was also true for flu
[23:18:24] <xedrIRC> So, this is the CTLA-4 protein? and we can bind this ligand anywhere?
[23:18:27] <@beta_helix> @susume And CO2 fixation as well, then... Can you post a feedback about that so we don't forget it?
[23:18:39] <Susume2> will do, beta
[23:18:54] <@EvdH> @CFC it really depends on the results if there will be a general strategy which is applicable for all the puzzles.
[23:19:31] <@Madde> Susume2: it is possible, just check "Show bonds (sidechain)" as well
[23:19:41] <@CFC> ...'suck it, and see'?
[23:20:00] <@beta_helix> that's the scientific process :-)
[23:20:08] <Susume2> thank you Madde, that works :-)
[23:20:13] <@CFC> [oh yes]
[23:20:23] <frood66> so our mission is to create a block - ok
[23:20:50] <@EvdH> @xedrIRC indeed this is the CTLA-4 protein. You are a little bit constrained in your search space. Actually you are constrained to the region near and around the natural binding area
[23:20:53] <@beta_helix> Thank you Madde!
[23:20:59] <@Madde> yw
[23:22:09] <@EvdH> @frood66 we'll in the application for T-cell cancer therapy one would define this indeed as a 'block'. But in general you can refer to it as a CTLA-4 binder
[23:23:51] <Chopsticks94> So are we trying to put the thingy in the protein?
[23:24:24] <@EvdH> @Chopsticks94 your making a valid point! That would be a nice strategy to go for :)
[23:27:06] <xedrIRC> hmm... do we need to be worried about burying this ligand too deeply?
[23:28:56] <@EvdH> xedrIRC: If you will buried it to deeply you will generate to much clashes. It think you should not worry much about a "too deep" or "too much on the surface" binding
[23:29:31] <Susume2> will the 3 residues we are designing be attached to a larger protein?
[23:29:44] <bertro> How confident are you of the shape of the big protein?
[23:30:03] <@rav3n_pl> small mutable? we can add more segments?
[23:30:41] <@EvdH> @Chopsticks94 thats true, but we know where the natural ligand binds so we already directed you in that direction with the constraints.
[23:30:42] <@beta_helix> @rav you cannot insert/delete, just mutate.
[23:31:18] <@EvdH> @Susume2 that will be eventually the goal, to graft these residues onto a larger scaffold which is able to keep your solution in the right conformation
[23:31:22] <@beta_helix> We're starting with 3 residues and we'll see what you come up with and take it from there!
[23:31:24] <Chopsticks94> Yeah, to give us an idea
[23:32:40] <@EvdH> @bertro this structure is extracted from a larger Xray structure which was solved to 3 angstrom resolution, so it should be relatively native :)
[[23:34:22] <@EvdH> As some of you may notice there is a nice cavity on the CTLA-4 protein, so that also might guide you into a nice solution
[23:35:20] <@rav3n_pl> this is cancer related thingy or i readed it wrong?
[23:37:26] <Susume2> that cavity will be challening to stick to - 4 prolines and a glycine :-)
[23:37:51] <@EvdH> @rav3n_pl yep this is definitely cancer related. As I pointed out in my blog post we can use solutions of the puzzle in several different ways. But one 'cancer related way' would be to prevent tumor cells from slowing down the immune system (which we obviously don't want, we want the immune system to kill the tumor cells)
[23:38:21] <@thomirc8> EvdH: and it applies to auto-immune diseases also?
[23:38:28] <@rav3n_pl> ah, understand. it is like hiv working too - supressing immuno system
[23:39:01] <@thomirc8> i.e. the grab-bag of diseases affecting the immune system?
[23:39:02] <bertro> Why only 3 seg for now?
[23:40:05] <@CFC> evdh - CD152 binding is a 'digital' on/off mechanism, or is it an analogue 'strength' binding that causes the effects?
[23:40:41] <@EvdH> @thomirc8 yep thats true. In an autoimmune disease the T-cell are on the other end of the activity spetrum; they are TOO active. So then we want to 'activate' the CTLA-4 pathway. We propose we can do that by fusing the CTLA-4 binder (which you are going to design :)) to a tag and in that way we hope to slow down a T-cell
[23:41:04] <@beta_helix> @berto 2 segments went really well for the Flu Puzzle, so we will post a small
[23:42:10] <@beta_helix> Ok... let's let Eric answer these final questions and then we'll move onto the chat with jflat06 and Seth Cooper... and I'll post 539 later today, ok? :-)
[23:42:15] <@EvdH> @CFC there are roughly 3 theories about the signal transduction of CTLA-4, it's not clear which pathway is the correct one. But you can imagine that if we can block the CTLA-4 by our design we create a more on/off fashion design
[23:42:21] <@thomirc8> EvdH: so the approach is to try to provide a "faucet" to regulate (control) the CTLA-4 pathway?
[23:43:44] <@CFC> thanks evdh
[23:43:59] <@EvdH> @thomirc8 yes :) actually there is another regulatory protein involved which, upon activiation, stimulates the T-cell. So if we are able to find binders for both CTLA-4 and this other regulatory protein (CD28) then we can really come up with a toolbox to regulate the T-cell's activity.
[23:44:02] <@rav3n_pl> it is posible that shithch we design will turn off more then we want...
[23:45:16] <@EvdH> @rav3n_pl we will never be sure on forehand, but by making a fusion of the design with a tag which is specific for a kind of cell we should be able to prevent a lot of unwanted side effects :)