The protein I have been studying has four cysteines which are predicted to form two disulfide bonds (one between the first and fourth cysteine and the other between the second and the third). However I am interested in solving the oligomeric state of this protein and want to know the significance of these cysteines. The protein is around 13kDa. When run in SDS-PAGE without dTT or BME, I see a band of around 25kDA. Hence I think it is a dimer. Is there any software that predicts and models the possible disulfide bridges between two molecules of a homodimer?
I have tried a couple of online softwares which require a PDB file to be uploaded but these do not recognize the PDB file I have from I-TASSER.
Hello Biraj,
maybe my answer is not what you expect, but I hope it helps.
First, which softwares did you already tried? so to know why could it be that your PDB is not being recognized.
Second, if you already have the 3D model, have you tested its quality? I'm asking just so the analysis you will do have more chances of being correct, and because maybe the reason why you can't upload them to the tools you mentioned you tried relies on the PDB lacking something.
Third, knowing you are trying to solve if these 4 cysteines are involved in the oligomeric state of the protein, you could do the analysis yourself before trying to use any software.
To do that you need to check some information about the protein. Is any of those cysteines on the surface of the protein? is there any homologue of this protein already crystallized so you can check if there are cystein pairs are forming intramolecular bridges? Are there enough homologues of the protein from different phyla so you can do an alignment an check if there are changes in the precense of these cysteines that can be related to the reported oligomeric state?
That can surely give you a better guess about what these residues are doing than a software trained to recognize who-knows-what in some training set with characteristics you don't know if your protein shares.
It seems like you want to characterize specific CYS residues related to the oligomerization not just determine the states. I am an NMR person, so I know what you can do if you can invest a little more money.
1. CYS CB chemical shift is extremely sensitive to disulfide bond formation. Backbone assignment needs to be done and you need to make 13C,15N double-labeled sample and collect at least two 3D data (CBCAcoNH, HNCACB). You can titrate your sample and see CB chemical shift changes.
2. You can observe CYS HG peak intensities in NOESY data. If chemical shifts are in the BMRB, you don't really need to do the assignment from scratch and just see CYS peak strips if HG NOE cross peak is fading and interdomain NOE cross peaks are appearing. It may require collecting 3D data and all about money and time. This is not a just prediction like 1 but the real evidence based on NOE. You can get 3D structure determined if you wish with that data which will require a lot of time and money investment.
I asnswer from a basic Biochemist point oif view. If your dimer is due to a S-S bond, the classical demonstration is to run a Western in (denaturing conditions) in both non-reducing and reducing conditions. If the dimer disappears in reducing conditions it is due to a S-S bond.
Rafael is right.
On an insilico level: open your i-tasser model in pymol and remove non-polar hydrogen. Resave and then try submitting to the prediction servers. try some protein protein interaction predictors such as ClusPro, GRAMXX and HADDOCK. You can then look at your predictions and see if disulfides are predicted. This is only useful because you want to solve the structure, otherwise the wet bench experiment would be even more compelling and sufficient on its own
- Badges
- Science method