Hello! Protparam instability index is used for In vivo stability of the proteins and is not related to thermal stability in solution. The chances of refolding after denaturation are decreased if: 1) you have a multi-subunit protein; 2) you have a certain number of cysteines forming disulfide bonds; 3) there are many ATP-dependent chaperones involved into maturation of your protein. This means that if its post-translational folding is ATP-dependent, its native structure is not the most stable one. So you cannot refold it to a native state

If you are expressing mammalian protein in bacterial system or membrane protein, there is a chance that it may go to inclusion bodies. Possibility of recovering such protein by denature purification followed by refolding is low but worth trying. It depends how big the protein is. If it is a multidomain protein can you truncate the unwanted domains ? A smaller protein is always easier to express and refold. If the protein is rich in cysteine bonds/glycosylation you may try to express them in mammalian/insect cells etc. Adding glycerol or arginine will not hamper your biophysical characterization in general.     

Regarding folding, the initial reaction is that it will be very difficult but without knowing details of your protein it is very difficult to know and suggestions that have been already made would be my response based on assumptions or guesses that may not apply. Even if it were to refold to the native state you would still have the stability issue as you currently face with purified native protein, right?  

Regarding biophysical analysis, again there is a need for details of what your analysis are.  The final formulation could influence the particular analysis such as field flow fractionation or differential scanning calorimetry, tryptophan positioning, titrations, etc.  In industry it is typical that a small amount of material be purified firstly for formulation studies before any significantly sized amount is made because it is of no use to make material unless one knows what formulation will stabilize it for further applications.  It may be better to invest time into partial factorial formulation studies rather than refold studies; unfortunately you'll need enough material but the findings may be publishable.

Hi shubbir,

My protein is non glycosylated having a 5 cystein residues. Its a human protein expressed in bacterial system. It has 6X-his tag at N-terminal. I am not removing that tag since it has a TEV protease site ( i have learnt that his tag does not effect the protein conformation studies). 

My question is if a protein by chance forms inclusion bodies then it should not get purified by the Ni-NTA chromatography, as the protein is aggregated. But I am getting a good protein yield after Ni-NTA. but it starts aggregating within a few hours. keeping at 4 deggree also doesnt help. 

Your further descriptions really help explain what is going on in the process.  Yes, the inclusion bodies would not be captured by the Ni column since they are particulate or insoluble in your current buffer.  It seems you never had native folded protein;  just what was captured from the soluble fraction by the Ni chrom.  If protein is still in imidazole elution buffer and/or at acidic pH then eventually there will be aggregation due to the cross-linking cysteine oxidation with incorrect secondary structure.  Yes, the protein will need to be refolded after the Ni step or even before chrom.  An odd number of cysteines indicates a covalently linked homo or hetero dimeric molecule or a monomer with a deeply buried unbridged cysteine or that unpaired cysteine conjugates to some other moiety.  If the molecular weight is under 50 Kd or so then you might have a chance to get it folded but it will require a partial factorial design and sufficient material. These refold designs usually have high, low, and medium concentrations of various components as well as varying the types of components.  In some cases I've reached as high as 100 different conditions.  Without violating company confidentiality/legal agreements, I can disclose that chaotropes, co-solvents, redox couples, pH's, salts, sometimes detergents and amino acids, are used in various combinations.  It is an art, but the shape of the band on SDS-PAGE (or peak shape from capillary electroph) can indicate correct folding in many cases for higher throughput anaysis.

If you do not want to take this route then perhaps eukaryotic cell culture expression can be taken where correctly folded protein with proper leader peptide processing is secreted for your Ni purification.  Since the protein has no NXS/T sequence then the major variants that might occur are aggregates,  O-glycosylation with variable occupancy, as well as proteolytic degradation from proteases such as furin.  

Do you know if the cysteines form disulphide bond  if so you will need a leader sequence and periplasmic expression. If disulphide bond formation is not part of nativley folded protein use high concentration of DTT to reduce formation of aberrant disulphide bond during purification. Just guessing that you have natively  folded protein that is gradually going in to aggregation due to disulphide bond formation during purification process. If so DTT will help. If your protein is rich in beta sheet, such protein have a tendency of forming time dependent aggregation in some cases even if stored at -20. I have worked with a beta sheet rich protein that had the same tendency of aggregation after purification in a day or so. To overcome that, just after purification I used to check activity and did the biochemical/biophysical analysis immediately and had to work for almost 18 hrs in stretch on those days. His tag is ok, technically it should not influence aggregation, bt you can not completely rule out that possibility.

@shubbir,

Its a monomer, but its crystal structure shows its as a homodimer in biological assembly. On SDS_PAGE only a single band is visible. I have used 0.5mM TCEP(reducing agent), it didnt work. Shall i increase the concentration or shall I use DTT instead. But does use of  DTT in elution buffer effect Ni beads in Ni-NTA column. Shall i also use B-mercaptehanol? kindly let me know the optimum concentration of both, so it does not hamper the purification as well as the structure of protein.

PDB structure shows 30% a-helix and 17% b-sheets.

The crystal data seems to confirm the earlier suspicion that the protein is multimeric and contains disulfide bonds which is the usual case for mammalian proteins having more than one cysteine.  Your preparation is currently monomeric because the protein hasn't been folded yet but once refolded do you think it would form a covalent dimer consistent with the x-ray crystallography data?  If there is crystal structure data then there's a tremendous amount of information available for this protein.  At a few angstom resolution the data will not only show the secondary, tertiary, and quaternary structural features but also will show which cysteines are paired. You should also know the activity assay for this protein if it is an enzyme and therefore know when it is properly folded.     

Remember the Eh (reduction potential) inside E. coli is very reductive and that is why your protein is expressed in the reduced state unless it can get to the outer membrane space and oxidize.  If it is an enzyme and relatively small with several disulfide bonds as questioned before, then you have a less than medium chance to correctly fold it even though the difficulty goes up tremendously for multimeric molecules.  This difficulty is raised even higher when the molecule is an enzyme because of the greater sensitivity of enzymes to conformational changes in the active site versus growth factors, cytokines, structural proteins, etc.  This is why it might be better to let the expression system fold the protein for you and therefore using expression systems that allow for chaperonins and oxidation to work could be better than trying to refold it.  

Since you have the His tag then take advantage of it especially because it is exposed on the protein's N-term which apparently is not involved in the active site conformation.  If you are concerned about any insoluble fraction after cell breakage then be sure to centrifuge well and filter the supernatant before Ni chrom.  You can't use reducing agents in Ni chrom because they reduce the Ni oxidation state and render it useless.  You can't alkylate the cysteines because that will cause irreversible modification and block the protein's cysteinyl disulfide formation. You might be able to lower the pH to around 6 so as to protonate the cysteine's sulfhydryl groups and significantly lower the oxidation (as aggregation) rate but crude E coli homogenates will likely start to precipitate (maybe with your protein) unless there's some chaotrope present but the tau nitrogen of the histidines will start getting protonated and not bind well to Ni.  You could try sulfitolysis to generate a reversible sulfite adduct of the cysteines.  How much time do you have for this project as that may dictate which strategy you want to take?

 By the way, if it is a non-covalent dimer then you will likely have at least one free sulhydryl group per subunit due to the odd number of cysteines.  This will present challenges when trying to stabilize the enzyme if this cysteine is surface exposed or binds cofactors or is a part of the catalytic mechanism.  You'll wish that it is deeply buried and not critical for structure of binding other moieties.  If the protein can tolerate it, a pH around 5 for the final purified molecule can be a good formulation since it will slow the oxidation of the cysteines through protonation of the SH group rather than relying upon reducing agents for mammalian enzymes that can end up getting their essential disulfides reduced as well.  Good luck!

If the crystal structure is available then you have tones of information. How different is your construct from the published structure ? I will advice just to follow the protocol that is available with the structure paper. TCEP does not hamper Ni-NTA purification. There is no standard protocol/condition for salt /TCEP/buffer pH. Try the possibilities available in the literatures. What will work for you is trial and error. As the biologically active protein is homodimer, I would assume it is domain swap homodimer. 

For what it is worth, we did a lot of work on acid denaturation of immunoglobulins. The dogma was that you regain combining site activity but lose complement activation potential when you neutralize acidified antibodies. We showed that if you re-nature Ig from pH 2 to neutrality and perform that in half pH steps and at low concentration, you regain complement activation potential. On the other hand if you do it quickly, dimers trimers as well as insoluble IGs with no complement fixing ability result.