There are buffer components that could interfere with binding, but I guess you have looked carefully at the protocols for his-tagged proteins and these components are not present in your extraction buffer.

I also guess that the sequence of your construct has been confirmed to contain the his tag.

Protease cleavage seems to be a possible explanation: what protease inhibitors are you using?

Regarding induction time: have you determined the kinetics of induction using PAGE and staining (or better, using an antibody, if you have one)?

Can you confirm that your protein is not present in inclusion bodies?

HI,

I agree with the comment on protease inhibitors...quite important to have the right amount and type in your mix...at what temperature are you doing your incubations? 4° or Room temperature.may make quite a difference. Also, using some extra (if you do not have them already) or a different detergent in your wash buffers may help. FInally, if your impurities are of low molecular weight (and sufficiently far away in molecular weight from your protein of interest), you may get rid of them by Ultrafiltration, Dialysis, etc.

In addition to the others answers, you already check that your protein is not in the inclusion bodies? Other possibility is that the his tag are not exposed on the protein surface, so are unable to bind to the column

In your case, you can add pre-cooled 1X PBS along with PMS and cocktail for digestion. You dont need to add imadzole here which is most probably eluting your protein at low concentration. After digestion, incubate your lysate with resin at 4C for 2hr with slowling shaking up and downward or you can also incubate at 4C overnight but then you need to wash the column well. After Incubation, wash the column with PBS for two times and then start with 10mM imidazole in PBS, 20,40,80,150, 200 and 250mM. you could get the purified protein with almost no impurities by this way. Good luck.

I have checked and am quite sure that the protein is not in the inclusion body. My protein with his tag can bind to the resin however it seems like the impurities can bind to the resin also...

I didn't add any of the protease inhibitors inside my mix. If I do, how much should I add? I have checked and saw that several kinds of protease inhibitor cocktails could be used during his-tag purification process. However, because of the larger volume of the lyse, does it mean that I have to add in like a lot? :(

By the way... the lysis process. I didn't add in any kind of a lysis buffer. All I do was suspend the pallet with buffer containing NaCl, phosphate and imidazole. Nothing else. And then sonicate the solution. 

Thanks all of you for answering.

To Zahir Shah:

I have tried to use buffer without any imidazole during digestion. However, my problem is not about my protein being eluted during low concentration. Hmmm I should say... it's the resin itself's characteristic to start eluting the target protein out during low concentration of imidazole. So it's not something that I can control... (This specific characteristic was listed on the product instruction "Please note:

Contrary to other available resins, bound His-tagged protein typically elutes from cOmplete His-Tag Purification Columns with a lower imidazole concentration, e.g., 25 to 45 mM.")

I did try a method similar to your suggestion earlier. I used no imidazole during digestion and then tried to wash the resin with 10mM, 15mM, 25mM, 30mM, 40mM, 50mM and then finally 250mM. Sadly, most of my target proteins were out during the 40mM and 50mM wash and not much was left when I wash it with 250mM :(

Nickel-NTA or TED coordination chemistry also affects binding of non-specific proteins containing 2-3 His residues in sequence. It would require careful elution preferably using a continuous conc. gradient of imidazole. PMSF may be tried as a protease inhibitor at 1-2 mM final conc. in the lysis buffer. 

Increasing the salt concentration in your binding and washing buffer will discourage unintentional binding.  If the buffer you are using already has salt, maybe try increasing it to 500mM.  

Maybe it has something to do with induction, if you suspect that your protein is being degraded and fragments containing the HIS-tag are being purified.  Induction at lower temps (25 C) might help to prevent degradation.  I start growing new cultures from an overnight culture, grow for 2 hours, then induce it and grow for two more hours at 25 C.  Pellet and then freeze for later.

Good luck to you. 

Re-reading your first post, it might be that you already have highly purified protein and you simply want to remove those smaller contaminants?? As they all have the same his-tag, it might be impossible to achieve that  by playing with the binding and elution conditions. However, we (and others) have used Sephacryl S-200 or  similar molecular exclusion columns for this.  Ultrafiltration might work, but the difference in size (and shape) between your protein and the contaminants is  critical (indeed, as it is for selection of MEC column!)

Based on what you have said so far, I think you should change Step 5. wash with buffer containing 30mM imidazole (pH8). Try doing the wash with only 10mM imidazole instead (you can increase this to maybe 15mM but I would start with 10mM first since it comes of with low concentrations). Then elute with buffer containing 250mM imidazole (pH8) on a gradient. (Are you doing a gradient elution or just a straight 250mM imidazole?) These two statements you made:

I tried using 25mM imidazole during the lyse process, however the protein just came right off when I tried to wash it for the first time. 

I also tried using 40mM imidazole during the second wash process, but the loss of the protein was a bit too much. 

tell me that you can NOT wash your column with with 30mM imidazole and still expect to have any protein bound. It is coming off at a very low concentration. I would not put any imidazole in your lysis buffer, not even 5mM. Because it is so easily removed with very little imidazole you are preventing it from even really being able to bind efficiently to the beads. Additionally, 25mM is just too much in your lysis buffer. We never add it to lysis buffer unless following someone else's protocol that has already been shown to be the best method. I think if you take the imidazole out of your lysis buffer and then decrease the concentration on your wash step to

If you have smaller bands than your protein lighting up on an Anti-His blot, then it is either your protein breaking down or non-specific bound proteins. I had a similar issue with one of my proteins and it turned out that I needed a special protease inhibitor, MG115. In all the literature I had seen on the purification of this protein only ONE had used it and I missed it the first time reading through all the papers. Upon having issues and further review of the literature I had read I found that one sentence where they say you have to use MG115 to prevent the protein from being degraded. Make sure your whatever protease or combination of proteases you use are in the lysis buffer.

If you have enough of your target protein, even with the possible breakdown of some due to degradation, you could take your sample and run it over a GF column that would be appropriate to get good resolution between your target protein and the smaller fragments. This may or may not be possible depending on the size of your target protein and the nearest fragment that is smaller than it in size. Ion exchange could be an additional option should you need to try yet another option.

Good luck!

Dear Geoff Margison,

Thanks for your reply. Yes, I already have a purified protein and the purity is around 80%. And I want to remove that 20% of impurities.  My protein is around 30kDa and the impurity bands are around 10-17kDa. I will definitely try using MEC.

Dear Derek Moates,

I have tried just washing it with 10mM imidazole first and then do the gradient elution. However, same results appeared. (Also, my protein sample become much more diluted.)

Dear James Carrillo,

I'm doing my induction process at 37 C right now. If my protein is being broken down inside the bacteria... I should just go with 25 C or even 16 C? I takes me 8 hours to reach OD600=0.6-0.7 and according to the protocol in our lab, the induction process was O/N. I know that normally they say that you should only do the induction for 2-4 hours... so should I retest the kinetics of the induction? or is lowering the induction temperature first is ok? Thanks so much!

Try it.  Add 2mL of O/N culture to 200mL of fresh media.  Shake at 37 C for two hours then drop the temp, induce, and shake for 2 more hours.  I haven't checked OD600 for a while, I just time it.  That is what I do and I am able to get little/no contaminants....but then the problem I have is getting my protein to work.

Are you growing your cultures at 37 until they reach OD and then cooling them down to 25 or 16. If you are inducing at 25 or 16 you need to wait until the cultures cool to the intended induction temp before adding IPTG. Most of the proteins we express require long O/N inductions of at least 16-18hrs at low temps. 

It seems to me that you can on fact calculate the abundance of your protein in the extracts. We have had up to 30% of total proteins but I would accept 10% as very reasonable. Tweaking the conditions for induction might not improve that by a great amount, and if the amounts you are currently recovering are acceptable, probably not worth the time and effort. I would focus on the MEC. I usually check the column's efficiency using commercial proteins of the appropriate size.

Do you have a resin without nickel that you could pre-incubate your lysate in (called clearing)? This might remove the nonspecific binding of your small protein contaminant. You then run the lysate that's already gone through the empty column (should now be free of proteins that bind resin/random things) through your nickel column to isolate the His tagged protein(s).

If you've done this then your contaminant probably binds nickel and I would do a size exclusion chromatography to isolate your protein of interest (since it's bigger than the contaminant). It also might be a protease problem as a lot of other people have suggested (use protease inhibitors).

Is there any way you could include a gel of your purification?  It would make diagnosis much easier!  In the absence of that, I'll take a few guesses:

1.  Induction:  If you have low levels of induction, you will get low levels of protein and higher levels of contaminants after Ni chromatography.  Also, extended induction at 37C (I think you said you do overnight) will possibly lead to increased proteolysis within the cell.  So the first thing to do is to optimize your induction conditions to give the optimal yield of full length, correctly folded, protein.

2.  Lysis:  Sonication can heat samples, leading to degradation of specific proteins.  Avoid whenever possible!  If you must sonicate, ensure short pulses and ample cooling time.  Also, DO add protease inhibitors (e.g. Roche complete minus EDTA) if you at all suspect proteolysis.  Avoid imidazole in your lysis buffer if your protein is eluting early from the column.

3.  Purification: why do you insist on using this resin if it says in the manual proteins come off at low imidazole?  I have never used that resin, and never will!  Although that said, some proteins will elute early from a decent resin (Like GE's for example) due to possibly the tag not being completely available/other factors that affect binding.

If you MUST use that resin, you could try loading it with a different metal ion in place of the nickel (e.g. copper, zinc, cobalt).  

Also note that there are E. coli proteins that will bind weakly to Ni resin (through His-patch or other means) that might contaminate your prep.   

So, make sure you only use enough resin to bind your protein of interest; apply in low imidazole; wash with enough very low (or even no) imidazole and then run a shallow gradient.  If your protein is too dilute after this, concentrate it or remove imidazole and reapply, bumping off with a step of imidazole.  

If it still isn't pure enough (and many will not be), you will have to run another column.  Size exclusion is usually a good second choice, as you can also use this to change buffers.  Ion exchange, hydrophobic interaction etc can also be used.  

Dear Derek Moates,

I tried yesterday... inducing the way you suggested. Hope the results will be better! Fingers crossed :)

Dear Geoff Margison,

Can I run my sample directly through MEC? Or do i have to do dialysis first and get rid of the imidazole (and also change the salt concentrations, currently the sample buffer contains 50mM NaH2PO4, 300mM NaCl and 250mM Imidazole) and then run it through MEC?

Dear Cole Peters,

I don't really understand what you meant by a resin without nickel that you could pre-incubate your lysate in... I have a lot to learn... Sorry :(

Dear Richard J Heath,

Thank you for your thorough answer. I have some questions though... I know people have been suggesting I use protease inhibitor during the lysis process. However, my sample (pallet+buffer) is 80mL per batch (and I usually purify two batches together, that makes it 160mL in total), does that mean I have to add in around 3 tablets of Roche complete protease inhibitor tablets every time? (because according to the instruction by Roche, every tablet is used for 50mL of lysate.)

Yes, add enough inhibitors (according to the directions).  

Just compare the cost of the tablets to doing your entire prep 2 or 3 or 4 times... isn't it better to spend a little more up front the first time and (hopefully!) have it work, than to have to do everything over and over again?  

Kelly, you seem to have followed the affinity method very well, so download the MEC instructions and do the same. You will learn that the sample volume to column volume ratio is critical and that you don't need to remove anything (the column will do that for you, and is one of its other advantages).  I strongly advise to run marker proteins first to ensure the column is OK.

And a note about proteases: we use pcomplete mini tables and add extra leupeptin, we also than add PMSF after sonication to break open the bugs. Despite these measures, we still get contaminants sticking to our affinity matrix!!

PS: DNaseI is 29.1 kDa and RNase A is 13.7 kDa. Your MEC needs to be able to separate these cleanly.

Ni columns have their limitations. Usually it's impossible to get it 95% pure after only the Ni column. I would suggest you do a ion exchange (Q or S depending on pI) and followed by a size exclusion. They will get your protein 95% pure, for most cases.