I scored the interactions in a protease-inhibitor complex using the 3D crystal structure, and I found these contacts involving two basic residues:
Arg18(NH2)-Arg61(NH1) [3,62A]
Lys8(CE)-Arg217A(NH2) [3,59A] and this is the closest one but there are 5 other contacts within 4A between these residues
Dear Rossana,
not sure if you only consider polar contacts, but for many enzyme-inhibitor complexes there might be more than just charge-charge interactions. In this case, maybe you are missing some water molecules to complete the pattern...
If not, you might consider the entity of buried surface upon ligand binding (the active site is normally filled with waters which are excluded when an inhibitor occupies the same space. This would cause an increase in entropy favouring the overall energetic balance, i.e. deltaG).
ITC measurements would tell you more on the energetics of binding, but if you base only on structural data, try estimating the amount of buried surface (delta-ASAs with Areaimol from CCP4 will tell you each inteactor's contribution).
Please note hydrophobic interactions might be underestimated from Thermofluor assays, hence a series of different techniques can provide you with contrasting data (or, better, can be affected by different types of bias).
And remember, the definitive incontrovertible data you have are the activity assays and the electron density... so everything must fit with them.
Hope this helps.
Ciao,
Marco
Dear Rossana,
not sure if you only consider polar contacts, but for many enzyme-inhibitor complexes there might be more than just charge-charge interactions. In this case, maybe you are missing some water molecules to complete the pattern...
If not, you might consider the entity of buried surface upon ligand binding (the active site is normally filled with waters which are excluded when an inhibitor occupies the same space. This would cause an increase in entropy favouring the overall energetic balance, i.e. deltaG).
ITC measurements would tell you more on the energetics of binding, but if you base only on structural data, try estimating the amount of buried surface (delta-ASAs with Areaimol from CCP4 will tell you each inteactor's contribution).
Please note hydrophobic interactions might be underestimated from Thermofluor assays, hence a series of different techniques can provide you with contrasting data (or, better, can be affected by different types of bias).
And remember, the definitive incontrovertible data you have are the activity assays and the electron density... so everything must fit with them.
Hope this helps.
Ciao,
Marco
You need to consider first of all the quality of the structure before making any conclusions. Is the electron density well defined in that area? You can use B-factors to assess the thermal motion of atoms. So if you have good electron density and low B-factors in that area, then you can probably make a conclusion that there are some repulsive interactions in the complex. If the electron density is poor, then no definitive conclusions can be made. High B-factors would indicate an area with a lot of thermal motion in the structure, but the deposited structure represents an averaged conformation which and this can introduce some artefacts as a result. There may not be any clashes of the atoms in the real dynamic structure, but in the averaged static structure, some atoms might be clashing with each other since their coordinates are only approximate.
Thank you for the answers. I will have a look into the B-factor since electron density is well defined in that area. It is possible also that I missed some waters during refinement although R factors were really Ok and then it was deposited.
I really appreciate this site for discussion since I can learn. Thank you very much!
Dear Rossana,
Obviously, Vladislav is correct that the fidelity of the structure is very important. Many times it happens that there is a small sequence shift that refines well, as some residues imitate others.
However, I wanted to raise another possibility. In many protein structures the notion of a local pH may play a role. A different protonation state of neighboring residues than expected can be detected. Very well described are strong hydrogen bonds (single proton shared with two acidic residues). I personally have seen a few cases of very close Arg or Lys that resulted from deprotonation of one of them.
Despite posing this possibility above I would rather look for errors in protein fit and unusual solvent peaks that would explain existence of negatively charged counter-ion not included in your model (it can be OH- or Cl-). You can easily discover it by lowering the contouring level of your maps. Well-refined models have interpretable maps down to 0.5 sigma.
Additional element of this puzzle is the directionality of hydrogen bonds. Arg is very polar in the plane of guanidinium group (forming strong and geometrically predictable H-bonds) but very hydrophobic with VdW packing in the plane to plane packing. This interaction is reminiscent of the DNA base stacking.
Some SPR studies could be performed:
Article Comparative study of Surface Plasmon Resonance, electrochemi...
- Badges
- Science topic
- Similar topics
- Biological Science
- Molecular Biology
- Biomolecules
- Proteins