Hi Rakesh. Very strange, but I'm more inclined to think that this is a protein precipitation problem rather than proteinase contamination. You could try adding 5mM EDTA to your loading buffer to exclude any complexes between His-tag and metal.

Some protease inhibitors do reversible inhibition of proteases..check how long these inhibitors can actively inhibit the proteases..it must be in the supporting literature. I suggest to add fresh protease inhibitor to protein before storage..hope it willl not degrade.

Is your phosphate buffer made fresh before each use, and do you sterilize it?

Have you tried to "monitor" how the protein disappears? eg. stored post-purification and taken samples for SDS-PAGE at certain time points to actually see that the protein vanishes and there really are no degradation products? Silver-staining of the gel might reveal hidden bands that are too small/weak to detect using coomassie. Have you tried storing a small batch of the protein under highly denaturing conditions (high urea, SDS, guanidium hydrochloride or guanidium thiocyanate - this should VERY well inhibit proteolysis) to check if the protein still behaves as observed? Could of course be that your protein is being eaten up by microbes (unlikely if you see protein disappearing at sub-zero temperatures). Anyway, sodium azide should delete this...

Let me get this straight. You boil a sample of the protein in SDS and load it on a gel and see a band. 5 hours later you load that same sample on a gel and see nothing?

@Steingrimur - that's what I understood and also something I've never come across before. Rather peculiar. The protein in the sample should be essentially denatured and as prone to degradation as any other protein. I have boiled proteins in loading buffer and left them on the table for several days and they've shown up just fine on gel.

I agree Steingrimur (wt!!!). Plus he's using a combination of protease inhibitor cocktails that have been shown to be effective in inhibiting protease activity in extracts of these cells when used in other expression experiments... also no detectable degradation products (presumably even at the gel front using silver). Hence my question about the sterility/ freshness of the buffers being used, because the loss has to be happening on ice and during transition to and from the frozen state (not at -80C, and most likely not even at -20C , although chemistry can defiinitely happen at this temp given a bit of time), and the mechanism of degradation needs to be significantly more resistant to denaturing conditions than most typical proteases (activity that would have been noted with extracts of these host cells long ago if it were present).

Whilst I undoubtedly agree with you Thomas, during my purifications I have never witnessed protein degradation on such a short term due to microbes. 5 hours in terms of protein stability is a very short time, especially considering the fact that the protein according to Rakesh is pure or at least very abundant and is being stored frozen or denatured in SDS-PAGE loading buffer.

@Rakesh, I suppose you have filtered your pure protein after purification? Say, 0.22 µm. This should remove the majority of microbes - at least most bacteria and fungi.

I agree Arne, but it has to be something that is unique to his system, since essentially the same system is used by many others without the observation of an abnormally robust, high activity protease activity (unless of course, the purified protein has protease activity and he is observing autoproteolysis).

Jump on in here anytime Rakesh... I think we all need more info about your system. :-)

@Thomas, indeed, autolysis could be one option, especially for intrinsically disordered proteins, although this should be quite diminished when the samples are frozen at very low temperatures. At least there should be a clear difference between samples stored differently. Yet I also think that autolysis should not occur easily in SDS loading buffer.

EDIT: yes, any additional details concerning the problem are most welcome.

Arne, I agree. What is described shouldn't happen at all, especially not as described. However, if we accept that it does happen, and it happens as described, then there is something unique to Rakesh's system that causes the observed result and one or more of the suggestions we've made is likely to lead to an answer... No matter how improbable, the possiblilities we've suggested should be assessed until new leads are established and an answer is forthcoming.

Is the protein a protease? If you concentrate a protease under conditions in which it is active (pH, salt) it can degrade itself. But if you boiled a protein sample in 1X final SDS PAGE loading buffer and ran 1/2 on SDS/PAGE and got a sharp protein band of the expected size, and then you ran the other 1/2 of the boiled protein sample +5 hours later and found no protein band(s), the only thing I can think of is protease K contamination of the SDS loading sample buffer. Pro K is active in SDS loading buffer, but after boiling even Pro K should not be active. Is the SDS loading loading buffer 1X after adding it to the protein sample? (i.e., are you using 4X SDS loading buffer). The only other question is, was the boiled protein sample in SDS loading buffer under your supervision the entire 5 hrs?

Thanks for all the discussion. Yes Steingrimur, it is the same sample boiled in loading dye and stored at +4 & -20 degree. Regarding the buffers, most of the buffers are freshly made or made from the stock solutions. I have repeated the experiment 4-6 times in a span of a month with same results. I have not done a silver staining, but have run couple of westerns for both His-tag and protein itself. It is gone after 5-6+ hours.

Protease inhibitors are added freshly every time as well as before storage.

I don't think there is any Pro K as it is purified on a Ni- column, and I can,t see a bond for the same on the gel. I use a 6x loading dye, dil. to 1x in sample boil for 10 min load half on gel and store half at +4 or -20 deg. Load it after 6 hours and it is gone. Protein disappearance from pure sample is faster as compared to boiled with LD.

Protein is a cytokine.

I think your protein may also have protease activity. Presumably at the high concentrations you have upon purification the protease inhibitors you've added are at too low of a concentration to inhibit, or they are not specific to the protease activity that may be associated with your protein. I would recommend doing what Arne originally suggested and do a time course for the degradation of your protein. Follow it with silver staining and blotting for His tagged peptides. If you detect cleavage products you should try to assess the cleavage sites by MS. Once obtained, you can use this information to select substrate proteins/peptides that either lack or possess the site for testing against your purified protein to assess it for activity. Also, if your protein does possess protease activity, you may be able to eliminate the autoproteolysis and stabilize the protein for analysis by substantially diluting the purified product in your final buffer containing the protease inhibitors (dilution series). Obviously this approach won't help you much with your cytokine experiments, but it may lead to new insights regarding previously unappreciated functions of your protein.

Also, a question. Do you use a reducing loading buffer, what's the reducing agent, and how fresh is the buffer?

I will try suggestion. Yes, I use reducing agent in the LD (b2M it is freshly added to the aliquote of LD I have to use).

Hi Rakesh. Very strange, but I'm more inclined to think that this is a protein precipitation problem rather than proteinase contamination. You could try adding 5mM EDTA to your loading buffer to exclude any complexes between His-tag and metal.

Are you purifying the protein from the inclusion bodies or the soluble fraction? If from IBs you can do several washes of the IBs with 1 M NaCl, 1 M urea, DNAse, RNAse to clean then up before you solubilize them in 6 M urea and bind the protein to the column, you can then do an on column re-folding by using a gradient to remove the 6 M urea, which should work fine for your size protein, and then elute with imidazole. When I have done that with BL21s I have not had a protease issue.

If you are getting the protein from the soluble fraction, then I think you really need to either: 1) dilute the eluted protein fraction out in buffer A (to reduce imidizole to a level that allows re-binding) and then re-load the protein onto the column and do a series of step gradients right up to where your protein comes off and then use a narrow step gradient to elute it (say only a 10-15% increase in imidazole), the protein sample will not be as concentrated but it may be free of the proteases; or 2) purify the protein immediately on another column to get rid of the protease(s) (cation/anion exchange). You will have to reduce the salt if your using it for the His-tag column, that can be done by dilution with rapid re-loading of the protein onto the second column or by dialysis (but may run into the degradation problem).

Did you find a solution for the problem?

is it protelysis? or precipitation?

Did you do desalting after Ni-NTA beads?

My idea is:

Addation of the protein to the surface of the tube, in which you store the protein.

You can do a western-procedure with the tube.

So if it is added to the surface you can see luminescence in the tube,

control shoud be a tube with the storing solution without the protein.

 Did you have imidazole in the protein samples? Imidazole can catalyse the cleavage of peptide bonds at high temperatures, for example when you boil your samples for SDS-PAGE.