I agree, the signal is too high and the band is too broad to say that it was only one band before the column. If you use fewer sample you will probably see the same result. Charlotte recommends a Western blot, is the image already a Western Blot? to me it seems like a western blot in which too much sample was loaded. It is very common your result. The first theory is truncated species, in which case, if it is a Ni-column you can try an imidazol elution gradient, with some luck you can have different affinity, which has actually happened to me. I imagine that you already standarized the expression conditions, if not you should try that, maybe there is a particular condition in which there is less degradation. Be careful to use the appropriate amount of sample, if you overload you won´t be able to see the differences. For other theories and suggestions more info is needed. If you are expressing in bacteria and the protein isn´t from bacteria you should analyze codon usage, considering that the degradation pattern is so specific.

I agree, the signal is too high and the band is too broad to say that it was only one band before the column. If you use fewer sample you will probably see the same result. Charlotte recommends a Western blot, is the image already a Western Blot? to me it seems like a western blot in which too much sample was loaded. It is very common your result. The first theory is truncated species, in which case, if it is a Ni-column you can try an imidazol elution gradient, with some luck you can have different affinity, which has actually happened to me. I imagine that you already standarized the expression conditions, if not you should try that, maybe there is a particular condition in which there is less degradation. Be careful to use the appropriate amount of sample, if you overload you won´t be able to see the differences. For other theories and suggestions more info is needed. If you are expressing in bacteria and the protein isn´t from bacteria you should analyze codon usage, considering that the degradation pattern is so specific.

Dear Charlotte and Carola,

Thank you very much for your insightful comments.

No, I have not done western blot at the moment. But that will be my next step in the coming few days. In the mean while, let me just show you my crude protein (before running on AKTA) electrophoresed on PAGE. Lane 7 and 8 depicts the soluble fraction with only single band. So from this image, do you still think that the other 2 ambiguous bands that I showed in the first image (the one with the red arrow) does not belong to my protein? I am just worried, if my rProtein is being degraded upon purification. 

Anyways, I will definitely enlighten you here once I am done with western blotting.

Thank you.

Dear Saranpal,

I have had similar problems with protein purification. I would suggest you to increase your washes to elute any non-specific proteins bound to the column. Another most probable reason could be that the proteins from your host strain are being pulled down by the Ni+2 column.

Dear Saranpal,

Could you describe briefly your protein expression and purification protocol?

Have your protein glycosylation sites?  

Dear Saranpal,

Once you run your western blot you'll know if they are degradation products of your protein or not.

If they are degradation products, then it would all come down to inhibiting the degradation. My guess is that you either don't add protease inhibitors to your lysis buffer or your buffer is sub-optimal for your protein's stability.

I harvest my cells (generally E. coli expression strains)  via centrifugation and discard the supernatant/growth medium. I then resuspend each cell pellet from a 1 L cell culture in 20 mM TRIS pH 8 + 500 mM NaCl + 20 mM imidazole + 1 tablet of EDTA-free protease inhibitor (Roche) to a final volume that is no larger than 30 ml for each pellet.

Working in a cold room at 4oC or at least keeping your samples on ice for most of the purification will also help reduce protein degradation.

The inhibitors will stop the endogenous proteases naturally found in E. coli from degrading your protein of interest when you lyse the cells and, since they are EDTA-free, they will not interfere with Ni-affinity purification.

Also, if your protein isn't properly charged it will become unstable and degrade. So, look at the pI of your protein and choose a pH for your buffer that is at least 1 point away from your protein's pI. However, the pH of the buffer needs to be 7.6 or higher for Ni-affinity. So, if your pI = 7 then go for pH = 8 to 9, if your pI = 8 the go for pH = 9, you can't go for pH = 7 as you protein won't bind to the nickel resin at that pH (the histidines in the His-tag are only charged and stick to nickel ions at pH 7.6 or higher).

If the bands you see are not degradation products then the next few points might help you.

Many proteins are naturally rich in histidines and therefore bind to the nickel resin almost as well as his-tagged proteins. Adding 20 mM imidazole to your lysis/binding buffer will reduce the nonspecific binding of these proteins to the resin and therefore increase the chances of your protein of interest binding to the resin instead. Use a gradient to elute your protein on an HPLC or, if you use batch methods or gravity columns, wash your resin with buffers containing 40 mM, 60 mM and 80 mM imidazole before eluting with 300 to 500 mM imidazole. This will give you a cleaner sample by removing more of the non-tagged proteins.

Adding 300 to 500 mM NaCl to the buffer also helps to reduce non-covalent binding of other proteins to your protein of interest.

I you need to choose a pH and salt concentration that are not optimal for your protein but necessary for the affinity purification step, then try to work as quickly as possible and change the buffer to optimal conditions as soon as you've eluted the protein from the resin.

I hope this helped.

I forgot, adding a reducing agent to your buffers might help stop your protein aggregating and other proteins from binding to your protein of interest via disulphide bridges.

However, reducing agents will interfere with nickel affinity chromatography as they will reduce the nickel ions on the resin. You can safely use 5 mM betamercaptoethanol on most resins and even up to 10 mM on HisTrap columns. Don't use DTT as it is a much stronger reducing agent and it will turn your nickel resin into a brown mess.

I would also reccomend gradient imidazole wash. Want is the downstream application this protein. How pure of a protein do you want. To me it looks like a very pure overloaded protein. You will always have minor contaminants but for most purposes that is fine. If you do need additional purity I would reccomend Ion exchange chromatography followed by gel filtration chromatography. If you beleive its indeed degrading (I dont personally think so) try to add protease inhibitors and keep it at 4c during all purification steps.

To me it seems like pure but over loaded protein.  In my case, I also used to find such multiple bands. I tried to reduce the load and the impurity band intensity reduced significantly.  You can also try a imidazole gradient as well as a second purification step to get rid of the impure bands. The other thing is, there might be some protein degradation during sample preparation step before gel run, esp. if you are heating the sample at 70 C for 10 mins or more. 

Dear Antonio Ariza, 

Your detailed answer is indeed useful. As you mentioned about pI of the proteins and pH of the binding buffer, what should be the pH if my protein of interest has pI of 4.2? I am using pH 8.0. I regenerated Ni-NTA column, but still I found that my protein is going to filtrate. 

Thanks,  

Sudhir, 

Saranpal , i do agree with Charlotte and Carola, WB will give some clear picture and as suggested by Alejandro Sevillano that might be beacuse of glycosylation if your protein has some sites. Have you done MS for all three band.

Dear Sudhir,

Make sure you include the full protein sequence for your pI calculation and that includes and tags. If your pI is 4.2, then a TRIS buffer with pH 7.6 to 8.0 should be ok for Ni-NTA. I would try to work as quickly as possible and then, after you elute the protein from the column, I would dialyse the protein into a buffer with pH 6.2 (MES, BIS-TRIS or PIPES will buffer at this pH).

After a bit of thought I think the best approach is probably to choose BIS-TRIS-propane as your buffer for all stages. You can use it for your lysis/Ni_NTA buffers and also afterwards, as it buffer between 6.3 and 9.5. So use this buffer with pH 7.6 to 8.0 for your lysis and Ni-NTA steps and then change it to pH 6.3 for your any other steps (ion exchange, gel filtration, etc...) you might need.

You say your protein goes to filtrate ... if you mean it doesn't stick to the Ni-NTA column then you might have a problem with the tag. His-tags can sometimes be occluded (especially if the protein forms multimers) so that they can only partially bind to the nickel resin. In which case I would change the His-tag of my construct from one terminus to the other to see if that helped (e.g. try a C-terminal His-tag instead of an N-terminal His-tag). Also, changing to a 10His-tag (these tags have 10 histidine residues instead of 6) will give you better binding as they have much better affinity for the nickel resin that 6His-tags (a strongly bound 6his-tagged protein will generally elute at 150 mM imidazole, a strongly bound 10his-tagged protein will elute at 300 mM imidazole, giving you a much cleaner sample as you can wash off more contaminants this way).

In addition with Antonio I usually add Guanidine at my loading buffer ( after loading, a refolding step is needed),

If you lose the sample in the filtrate you have 2 possible troubles, 1) All Ni sites are binding with imidazole/ residual proteins , you should prepare a fresh Ni-NTA column again. 2) I agree with Antonio, you should to adjust the loading ( sample) buffer to 8.0-9.0.

Hi. If your expression was in E coli, please note that the 70kDa E coli chaperone called DnaK often gets co-purified with His-tagged proteins. That should, (if that's really the case) explain the band above 66kDa mark and below your protein. You can do a gelfiltration of your eluted protein (without cleaving tag) in, let's say Superdex75, and check where your protein elutes. It may well be that the 70kDa protein might get separated from your protein. Secondly, if that is really the 70kDa chaperone, you can include a wash step with apprx 10mM ATP during your Ni-NTA purification. Since DnaK is an ATP dependent chaperone, in principle, you should be able to wash it out before eluting your protein. I haven't done it myself but I have read that it is possible.

Hi. Try to lyse cells in a low salt tris buffer pH= 8.0 and after binding of lysate supernatant to column was with high salt low imidazole ( 20mM) tris buffer ph= 8.0. This is critical as this step removes most of the unbound protein. Elute your protein with high concentration of imidazole approx 250mM. If still you have contaminants try doing AKTA (FPLC) or GFC to get a single band of interest. But make sure you have enough protein to start with before doing FPLC as a lot is lost in the process. All the best. Hope it helps.

Ensure there is enough reductant to completely reduce your protein on the  gel (sometimes a remaining internal disulphide will produce species with smaller apparent masses). If the bands persist, I would separate the bands better, maybe by running the gel longer or by using a different gel. And then try to get MS identification of the species using in-gel digestion.  This will tell you whether truncations of your protein have occurred or you have co-purified a nickel binding protein.  The iD will help you plan future purifications.

It looks like you already have some good suggestions here!

1. It is not uncommon to have copurifying proteins from a nickel column.  I have often tried to overload my column to compete away other proteins that bind less tightly to get more pure protein even though some of it is lost.  I would also suggest using a second purification step - I have used gel filtration, but you could try ion exchange too.  Also, I have sometimes found it useful to purify with a cleavable his tag - first purify with a nickel column, remove his tag, then put it back over a nickel column and the cleaved protein should be in the flow through (hopefully any contaminating proteins stick to the column again).

2. As others mentioned, misfolding or aggregation could be a reason that the His tag isn't binding and/or that you have a chaperone binding.  Sometimes inducing at lower temperatures (like 16C) for a longer time can help  to reduce this.  I tried once adding different amounts of IPTG (I was using a T7lac promoter) at different temperatures (37,30,25,16) and then used ultracentrifugation to look for the fraction with the most soluble protein.  As others have mentioned, an ATP wash can also be used to get rid of chaperones.

What did your Western show?

Dear all,

First of all thank you very much for your critical comments pertaining my query and secondly sorry that I could not revert to your comments ASAP during the weekday. Now that its is the end of the weekend, hence I am answering those questions imposed to me:

@ Srikanth Sir: Actually I have washed with 3-4 column volume of washing buffer (15-20mL) supplemented with 20 mM-50mM Imidazole. The protein elutes at 200mM of Imidazole

@ Alejandro Sir: Briefly, protein is expressed in 1 liter of LB culture at 25°C using 0.1mM of IPTG (I have tried 6 hours and overnight expression, but the results are the same), pelleted and sonicated in 50mM Tris supplemented with 100mM NaCl (pH8.0). The resulting supernatant is filtered (total volume ~30mL) and pass-through the Ni-Nta column. Of course, the column is first equilibrated with 50mM Tris supplemented with 100mM NaCl (pH8.0). Before elution, the column is washed extensively with 3-4 column volume of washing buffer supplemented with 20 mM-50mM Imidazole and my protein elutes at about ~200mM of Imidazole. My protein is a non-glycosylated protein (Not sure if the tag is? In this case, do I need to consider the glycosylation site for the tag? I am using pET43.1a Nus A tag).

@Antonia Sir: Noted, will definitely try increasing NaCl concentration. Thank you for that lengthy comment ☺

@Sreejesh Sir: So you don’t think it is a truncated version of protein? Oh, that’s good.. Recently, a colleague of mine has also used the similar tag, but she often gets a single species after purification. We are using the E. coli strain (BL21DE3).

@Sujit Sir: Actually, I have also tried using the imidazole gradient and have also performed 2nd purification step (with Ni column again) as advised by my PI.

@Inderjeet Kaur: For the WB, do I need to raise an antibody specific to my protein or can I use the anti-His antibody? No, as mentioned above, I do not have any glycosylation sites in my protein at least (but I have never checked if my tag has it?) Do I need to consider the glycosylation site in my tag? I am using pET43.1a (NusA Tag) from Novagen.

@Jyotishman Sir: I did not know that the chaperone has an affinity to the Ni column. But if this is the case, then why isn’t my colleague facing the similar problem? But anyhow, I will definitely try gel-filtration or perhaps the addition of ATP to get rid of that chaperone. Perhaps, I will first try playing with the NaCl concentration, just as advised by Antonia sir and the others. Thank you for enlightening me about the chaperone.

@Ankitta sir/mdm: Indeed, I have used FPLC system. Yes, I will try using high salt concentration.

@Gordon Sir: Yes, I will try troubleshooting the problem step-by step and will also include ~5mM B-Me in my binding buffer and see what happens. Particularly, my protein does not have any cysteine residues, but the tag does have some. Will also attempt to perform MS on that 3 species. Thank you.

@Amanda mdm: My protein is expressed at 25°C using 0.1mM of IPTG for 6 hours. Have tried previously expressing it at 15°C overnight (~16hours), but I get very less protein. Ya, maybe I should try expressing it longer at this particular temperature as this might lower the chances of protease inhibiting/truncating my protein. Will definitely keep in mind. Thank you.

@Fedora Sir: I have NOT tried western blot at the moment. For this case, do I need to raise a specific antibody for my protein, or can I use the anti-His antibody to see if there is any degradation/truncation?

Another question to all: For this case, do I need to raise a specific antibody for my protein, or can I use the anti-His antibody to see if there is any degradation/truncation?

Once again, thank you very much for taking out some time commenting on my problem.

~Saran~

Hi You can check HIS tag antibody too for WB and secondary antimouse 1:5000

But it is always good to have specific antibody:)

The anti-His antibody should bind your protein with a His tag, but it will also non-specifically react with other proteins that have His in them.  Make sure you add a negative control lane (no induction will have some background that can be significant unless you're using pLysS, you can also have some straight up BL21DE3) to get an idea of what bands might be nonspecific.  You should try it as a first step, but it is certainly easier to blot with an antibody specific to our protein.

Also I didn't see any protease inhibitors in your lysis buffer (although maybe you just didn't mention it) - the roche EDTA-free complete tablets for Ni-NTA columns are nice and easy to use.