 |
 |
Thank you for all the advice. A few people have said to include the peptide in the cryo, we did not do this so we will certainly do so next time. Also good point Artem about the physiological pH! We do know the affinity of the peptide, 1.5uM by ITC, but we will try higher peptide concentrations as suggested to try and saturate the sites. There is a possibility the crystal packing is obscuring the binding site, although we are co-purifying the peptide and protein to give them as much time together as possible before crystallisation. Also I really hope we don't have to, but good point about changing the peptide length for crystallisation, we will go back to this if all else fails! Thanks again, Nicola -----Original Message----- From: CCP4 bulletin board [mailto:[log in to unmask]] On Behalf Of DUMAS Philippe (VIE) Sent: 03 April 2016 11:57 To: [log in to unmask] Subject: Re: [ccp4bb] Peptide co-crystal advice? Le Samedi 2 Avril 2016 19:15 CEST, Gyanendra Kumar <[log in to unmask]> a écrit: Hello We made exacly the same kind of co-crystallization protein+peptide. We noticed that the intermolecular interactions in the packing could be strong enough to prevent occupying all binding sites (Burnouf et al.,J Mol Biol. 335(2004)1187) Also, practically, E. Ennifar introduced ITC as a very efficient tool to prepare the material for crystallization: instead of letting the titration go to completion (i.e. reaching a stoichiometric ratio significantly above 1), you stop it just above 1 and retrieve the materiall in the cell for concentration and crystallization [Da Veiga et al., "Isothermal Titration Calorimetry: Assisted Crystallization of RNA-Ligand Complexes." Methods Mol Biol. 2016;1320:127-43. doi: 10.1007/978-1-4939-2763-0_9] This allows one to get the minimum and sufficient amount of ligand. Also, analytically, this reveals whether 100 % of the macromolecule can bind the ligand (if the concentrations are correct). If significantly less than 100 % can bind the ligand, then you know that all attempts of crystallization will lead to a mixture of molecular species and, likely, to poor crystals or no crystals. At the extreme limit, you may well discover that your macromolecule/ligand pair do not interact significantly. I hop this will be useful. Philippe Dumas > Hi Nicola, > > I would suggest that you keep increasing the concentration of the > peptide as long as it results in crystals appearing in your drop. If > your peptide has low affinity for the protein, it may take a lot of it > to see it in the crystal structure. Try 10, 20, 50, 100 fold higher > concentration of the peptide. Also, shoot your 1:5 ratio crystals, it > may already be there. In some cases, if the crystals have large enough > solvent channels, the morphology of the crystals may not change > significantly upon binding of the peptide. > Congratulations on having a fast crystallizing protein at hand. That > makes it much easier to try new things and test your hypothesis at a faster pace. > Also, you can try variation in length of your peptide, if increasing > the peptide concentration doesn't give you the complex. May be, > addition or deletion of a residue or two will result in better binding > and you will have your complex. > > with best wishes, > Gyan > > > On Fri, Apr 1, 2016 at 10:58 AM, Nicola Evans > <[log in to unmask]> > wrote: > > > Hello all, I have a question regarding co-crystals with peptides, > > please do share any experiences you may have had. In my group we are > > trying to crystalise 15KDa protein with a ~1KDa peptide (related to > > a binding partner). We initially set up apo crystallisation and > > protein:peptide at a > > 1:1 molar ratio. We had crystals in both with very similar > > morphology (plates), the protein:peptide mix crystals diffracted > > well and gave us a good structure, but no peptide was bound, however > > the purely apo tray gave crystals with much poorer anomalous > > diffraction. It looks therefore like the peptide is helping the > > protein to crystalise in a more ordered form, but maybe the > > concentration isn't high enough to reach saturation. ITC indicates that this peptide does bind to our protein. > > > > We have set up further crystallisation with a protein:peptide ratio > > of 1:5 and have again grown crystals, however the morphology and > > plate profiles look exactly the same as before. In my experience > > with compounds this is quite common (and of course plates are a > > common morphology) but would you expect with the larger peptides to > > see this? Or would you expect something different? If not in > > morphology in the cell dimensions of the crystal? Is it worth > > collecting a dataset for these crystals to see if the peptide is bound? > > > > Some more information is that the crystallisation is very fast > > (concentration dependant), within minutes to hours. The peptide is > > also co-purified with the protein. > > > > Thanks in advance for any help. > > > > > > -- > Gyanendra Kumar, PhD > St. Jude Children's Research Hospital, Department of Structural > Biology, 262, Danny Thomas Place, MS-311 Memphis, TN 38105 > Phone: 901-595-3839