Hey Paul,

There is no EDTA at all. Since , i am trying to purify a membrane protein complex from mitochondria , the buffer has 0.009% DDM ,150mM KCl and KPO4 pH 8.0. The washes have 5 and 10mM Imidazole, while the elutions range from 50, 100 and 200mM imidazole. I used Qiagen Ni-NTA agarose.

Depending on the protein folding the HisTaq might not be accessible, means not on the protein surface. To find out, if this is your problem, you only have to purify the denaturated protein.

Adding to Marlen's comment, have you tried expressing the protein with the His-tag on the opposite end (to which it is now) of the protein? In other words, if you have it on the C-terminal end of the protein try using it on the N-terminal end. That has worked for my before.

if accessibility of your his-tag is a problem, you could try expressing it on the other terminus of your protein maybe. Do you need a his-tag though? Interesting that without a his-tag your subunit takes 50mM of imidazole to elute- are you expecting histidines or any other potenital nickel binders on the surface? good luck

I would guess the HisTag isn't there -- imidazol concentrations of 50 mM shouldnt elute the protein. Maybe move the Tag to another position (even internal Tags work quite well, or place a linker in front of the His-Tag) or try to use another purification strategy.

Hi Murugappan, So it looks like there are a couple theories out there that you can rule out with some control experiments. One is that for whatever reason, you really don't have a His tag (proteolysis, cloning mistake, etc). Agree with Sezer on that count. Do you have an antibody for both your membrane protein segment and the His tag? The folding issue is a possibility but it is a bit of a long shot as proteins like to fold their hydrophobic stuff inside- not the very hydrophilic/soluble His tag. Marien is not wrong but Occam's razor says control for the simple explanation first. Also, we need more information. You say you are purifying a "complex". Please explain. Another control I can suggest is to run a positive control His-tagged protein through your column elution protocol. Make sure the capture and elution protocols are working.

There are a few explanations/comments that I can think of:

1. your protein does not contain a hist tag: check seq to make sure

2. Hist tag may be sterically hindered by protein folding etc and therefore does not have access to Ni or has lower affinity

3. Check for non specific interactions by increasing salt to 0.5-1M NaCl, include Tween 20 in buffer (0.1% or less) or include 10-20mM imidazole in loading buffer

4. most proteins contain free histidine and other amino acids that have affinity for a metal ion column (cysteines, tryptophans) and therefore will bind the resin. Increasing stringency usually gets rid of these interactions because they are of lower affinity.

5. Try doing a gradient elution if possible instead of step elution you may find your proteins do elute slightly differently over a narrower range.

good luck

1) At times pH plays a very important role. Can you tell me what is the pH of ur binding, wash and elution buffers? u can try different pH.

2) One more thing u can do is that u can try NaCl in your buffer and try eluting with increasing concentrations of NaCl.

3) Also, keep in mind that some proteins bind less efficiently in the presence of Imidazole. So try keeping imidazole to 1-2mM in ur binding buffer and abot 20 mM in wash buffer. Also try range of Imidazole (40, 60, 80, 100, 120, 140-----300 mM Imidazole) for eluting so that u can standardize the concentration of Imidazole at which protein elutes. Or u can even omit Imidazole from binding buffer. But first try with 1-2mM of Imidazole.

4) If nothing works, U can purify your protein in denatured form in the presence of Urea or Guanidinum chloride. But sometimes proteins loose their activity upon denaturation. Then, u can try renaturation of your protein on column itself.

Even I faced the same problem.

Wish U Good Luck.

if the hist-tag in the extreme ( N-ter or C-ter) of the protein is not accessible,then, you may try to clone the hist-tag in the other extreme of the gene,and it is probably that the hist-tag in that extreme be more accesible to bind to the resine.

Hi! Are you using a new column or it had been used before ? If so, make sure that the column is properly regenerated since it might loose its binding capacity after few purifications. I've seen in some occasions that columns which don't retain properly His-tagged proteins they do so once the matrix is recharged with Ni. It will also allow you to get rid of any contaminants that will bind the column at low concentration of imidazol. Good luck!

Binding is highly pH-dependent. Is the pH in your binding buffer OK ?

As said below, check pH of binding condition. Imidazole is a tricky buffer.

I do agree with Fritz Boege the imidazol may change pH of your buffer, and as other suggested try to put His-Tag on different end of your protein it is quite possible that this tag is hidden by other proteins in your complex.

One question here. Is this in an E.coli expression system or mammalian or baculovirus. If it is not accessibility or a pH issue and its in one of eukaryotic systems the media contains divalent ions that prevent/reduce binding. In many cases this can be overcome by Am2SO4 precipitation and resuspension in a binding buffer, or by dialysis. If this is the problem it is possible you could try the New Ni affinity resin from Roche since it is stable to reducing agent and EDTA. We have tried it and those claims are correct. However, this comes at a cost of reduced affinity and proteins elute at much much lower NiNTA. This in itself can be a positive if you are having trouble with oxidation or you need lower non-specific binding (which does occur) in a pull-down this issue should be weighed carefully if you test this resin.

Hi Everyone,

Thanks for your response. So a bit more information on my current issue.

So i use yeast as the model system and inserted the linker (GARGS) and 6X His tag onto the C terminus of COX 13 ( subunit of complex IV of respiratory chain complex) through homologous recombination. Sequencing results came out clean with the linker and His tag on to it. Isolated the mitochondria from yeast and tried to purify the entire complex IV through a His tag on one of its subunits. I do have the antibody for other subunit (COX 4) of complex IV. I tried using mouse anti his antibody too but its not so clear on the Western Blots so far. There is a diffused band around the same molecular weight but not super clear. I tried using different salt concentrations (like 300mM NaCl) but in that case it elutes out at 10mM Imi. So far, i know at DNA level we have got the His Tag, but there is no proof it is present at protein level. This purification is possible and people have shown it before. The link to the reference is here

http://www.ncbi.nlm.nih.gov/pubmed/22394221

Hi Priya Arora,

The pH of all the wash and elution buffers are 8.0 . I do add 5mM Imi to the binding buffer. Since i am trying to purify the entire protein complex through one of its subunits , denaturing it ll defeat the entire purpose. I need to purify it in its native active form.

Agree with all the comments above. Here are my 2 cents on other issues:

1. Sounds like you might not have a good anti-his tag antibody, if you are using RGS-6his sequence, are you using anti RGS -his antibody? I used RGS-his mAb from Stratagene many years ago and it is very good antibody, other anti-his antibody did not work well.

2. Do you have to use Saccharomyces cerevisiae in you case? It is not a very good expression system (Pichia pastoris will be a better choice). We have switched to E.coli system years ago and purified more than hundreds of functional human, yeast proteins, including protein complex (rpa, rfc) and mitochondria proteins (mtRNAP, mtTFB2, mtTBA, mtpol Gamma). Novagen sells vectors that you can purify 3-5 proteins from same culture, which is very useful to purify protein complex. Another option is to use Sumo ULP1-his tag system (in E.coli expression system), we noticed that this system not only helps protein expression and solubility (may not work to some proteins), but also the double tags can be cleaved very efficiently/specifically by ULP1 Sumo protease (comparing to Thrombin,TEV, and EK), only need 1ul (12U)/mg protein for nearly 100% cleavage at 4*C overnight.

Hi Michelle,

I used Qiagen Anti-His Antibody for western blots. Yes, our lab use Saccharomyces cerevisiae as model system. E.coli system looks like a good choice and i ll look into that. Thanks for your suggestions.

you are welcome! Please let me know if you want to try Sumo-his tag system, you will get free Sumo-his expression vector when using our ULP1 sumo protease.

[email protected]

Are you adding detergent since Cytochrome c oxidase is a membrane protein. Otherwise you might get aggregation of the protein instead of binding

Hey Jenny,

Yes i add 2% DDM to solubilize it initially and the buffers contain 0.009% DDM.

Have you checked that your protein is i) expressed and ii) localized to mitochondrial fraction? Of course, you'll need an anti-His antibody that works. Your problem might be not in the purification procedure at all.

If you have his tag sequence in DNA it does not mean that you will have it in the protein. Sometimes Pichia pastoris expressed proteins have no his tag (I don`t know if it is expressed without it or it is cleaved. My guess it is the 2nd option). It happens when using pPICZalpha vectors. I have no experience with other yeast systems but I also saw the protein without His tag expressed in S2 cells. I never heard the good explanation and solution of this problem.

It seems to me that either the protein complex is preventing your protein from binding to the Ni NTA or that the his tag is not there.

So to rule out the first case, I would add urea to 6M-8M and see if your one protein sticks under these conditions. If it does, it means you will likely have to change the tag, or try a Ni resin that is not as bulky. I like the Resyn Biosciences magnetic beads as an alternative. If this does not work, then you need to change the tag, either by adding a linker, or changing the tag to MBP or GST.

I think that there have been many useful suggestions so far; however, I would like to add some of my own. I'm currently purifying a multispanning membrane-protein with a his-tag in its N-terminus and I have observed several interesting things: for example, adding 3 glycines between the his tag and the rest of the protein has been helpful to me because glycines act as a molecular hinge and makes the his domain more flexible thus facilitating the interaction with the resin. Another important thing is the yeast strain you use: I'm working with a W303 strain right now with good results but, previously, I tried a BJ5457 strain because it lacks vacuolar proteases and the protein didn't stick to the resin (?). It seems as if the his-tag did not appear in the protein expressed by the latter. Another point worth of consideration is checking (if you have structural models) whether the his-tag might occluded and, consequently, unavailable for interaction with the resin. Finally, in my experience, it's better to incubate the solubilised protein with the resin in batch and then pack the mixture into the column

Hi Jose,

Those are really useful suggestions. I did use W 303 strain and currently i am trying to use 6M urea to denature the protein complex and see if i have got the His tag. I chose this subunit 's C-terminus end based on structure model and it has been shown in the literature that this should work.