Information on proteins size would be useful. If both are large, than the line broadening is expected. You should record 1H-15N TROSY on the complex to see if signals recover. That will tell you whether fast relaxation or chemical exchange is killing the signal in regular HSQC. If relaxation is the problem, than you have chance for the assignment by recording better experiments ( TROSY, 13C detected experiments).

Dear Nenad Juranić ,

Thanking you for your Reply . We record the TROSY -HSQC for protein complex ( 21kd (A) + 13kd(B) = 44 Kd Total size of the protein complex ) . We did not observe the significant difference in case of TROSY -HSQC Experiments .

you observe intermediate exchange behaviour of your signal upon complex formation what in also expected for a Kd~6 uM. As chemical shifts in the complex change especially for the binding interface, assignment based on the chemical shifts of the free proteins is not possible. How many intermolecular NOEs did you observe?

Would they be sufficient to calculate a complex structure?

It sounds like spin-labelling and determining PRE would be the best option to obtain experimental information about the complex structure. In that case you are not restricted to the binding interface.

Dear Sigrun Rumpel

Thanking for you comments . We are encountering line broadening upon complex formation . PRE also can cause the Line broadening . Do we think it would be the best option ? . I am afraid that we may lose all the information . But I am not sure about that . I observed 9 intermolecular NOE between them with complex and used for HADDOCK .But Validity of the NOE on the basis of free assignment is not acceptable for publication . So far I did not find any literature to support my data in case of line broadening .

Do you observe line-broadening of all residues or only at the interface? For PRE, you have to calculate the distance from the line broadening but the spin-labelled residue would not be bound to a residue of the binding interface. 9 intermolecular NOEs sounds oK if they are spread across the interface. If you are certain about your assignments you can use them.

Dear Sigrun Rumpel ,

We did sequential N15-HSQC titration ( 1:0.2,0.4,0.6,0.8,1.0) . we Observed line-broadening only at interface ( 1:1 ) , But if we tried to over saturate the binding partner say ( 1: 2 or 3 or 4 ratio ) , Spectrum becomes more worse and caused line broaden over complete spectra . Because of limited solubility of the protein complex and stability issue and we observed precipitation of the protein .

Specific labeling of suspected residues in the binding region would afford assignation.

How ambiguous are your assignments? It's difficult to tell if you are encountering actual ambiguity (multiple plausible assignments exist) or another type ambiguity (only a single assignment is probable, but cannot be definitely established without reference to outside information).

In other words, is there another binding site consistent with your data that is plausible based on the structures of the free monomers?

Do you have other outside information such as mutation studies to back your interface model?

Thanks to Nenad Juranić and Jeffrey Brender for your reply ,

Interface Model is consistent with Mutagensis data (ITC) , we Mutated 2 electrostatic residues and few hydrophobics of protein A, The binding affinity reduced 3.6 fold and 150 fold for electrostatics . for the variant which triple mutated Hydrophobic Residues, Binding is not detectable with target partner from ITC . Few residues of binding interface for protein B are conserved with residues for other binding partners ( Published data ).

At this point then I would consider the opportunity cost of completing the full assignment in terms of other research that is not being performed while you try to rule out unlikely possibilities.

Thank you Jeffrey Brender for your comments !