Insoluble Proteins - Low Ni-NTA affinity and protein crashing

8 posts / 0 new
Last post
louhazosc's picture
Insoluble Proteins - Low Ni-NTA affinity and protein crashing

working on a series of insoluble recombinant proteins with a 6xHIS tag (N-term). Solubility analysis has been done and pellet MUST be dissolved in 6M GuHCL.
Following pellet re-suspension for one hour, it is applied to Ni-NTA and subsequently washed in a buffer series containing 20mM Tris-HCl/250mM NaCl/6M GuHCl/1mM Beta-Mercaptoethanol and increasing concentrations of imidazole ranging from 5mM to 500mM.

Protein has found to elute at 20mM Imidizole.
There is the first issue - SEEMS like a low affinity to Ni-NTA.
FYI - The isoelectric point of the protein is 5.7ish and buffers are at 7.5-8.0

Then there is the next and MUCH bigger issue.
Proteins are dialyzed against 20mM Tris, 150mM NaCl, 0.3M Arginie overnight and upon coming in in the morning there is a pretty good amount of precipitate right in the dialysis casette. Spun it down and looked at the supernatant and the precipitate pellet on SDS-PAGE and yep, the protein is in the precipitate.

Now one big thing is that this is a protein fragment so I'm not sure that re-folding is a huge issue.

What I am in the middle of right now is that I loaded the 6M GuHCl dissolved pellet onto Ni-NTA for an hour. Then, ON COLUMN I am running a 6M to 0M UREA gradient (in 20mM Tris,250 NaCl, 5mM Imidazole + Urea) and THEN doing my washes and imidazole elutions without GuHCl present. Going to run those out on a gel soon and see if it still crashed on column.

So if anyone has any ideas about the low affinity and more importantly if anyone knows of a magic buffer I can dialyze in that would be so appreciated. I have to try and keep it somewhat physiological, detergents are ok. I many nay people have had this issue when you try to get things into a physiological buffer from Urea or GuHCl.
Speaking of, I have also tried a TBS +0.5% Tween with a hint of EDTA (say .5mM) dialysis buffer and not good. This buffer was used afer column elution.
Also, I am about to try tonight...Dilluting the protein out to a very low concentration , I have no idea what it is now but still, and do a dialysis with that. I can only say that after running it out on a gel there is a good deal of protein in there so possibly dillution might help.

I know 6M GuHCl is common on Ni-NTA so I don't know. I would like to do a series of micro-dialysis screens but we have neither the supplies or time at this point.


kjosephs's picture
I too am puzzled why the

I too am puzzled why the protein would elute in such a low concentration of imidazole.  Under denaturing conditions you have avoided problems with display of the his-tag to the Ni, so the His6 Ni chelation should be pretty tight.  Is your protein flowing through or just binding weekly? Did you adjust the pH back to ~8.0 after adding 6 M guanidine-HCl to the buffers?  
Many proteins express really well but are entirely insoluble and the washed inclusion bodies are already 70-90% pure, so you might just skip the Ni step all together if it is not enriching your protein that much.
I guess if it is coming off reasonably pure at 20 mM then you can focus on your  MUCH bigger problem--refolding.  
Refolding efficiencies (especially if not optimized) can be really low, is there any soluble protein after dialysis on SDS-PAGE?  Do you have an acitvity assay to check the supernatant with?
If your protein does not have disulfide bonds,  then you should keep reducing agent (like 5-10 mM DTT or BME) in the refolding buffer to avoid disulfide linked oligomers.  If you have disulfides, then you need to oxidize them by including oxidized glutathione in the refolding buffer to soak up the BME you included in the solubilization buffer and to form disulfides, you can start with 5-10 mM.  If you have a few disulfides, then the efficiency of the refolding and oxidization to the native form could be really really low.  I have had to accept yeilds of 1-5% and got what I needed by brute force.  Even for proteins that refold efficiently there is always some precipitate, so focus on what is in the supernatant rather than assuming that what looks like "a lot of precipitate" is a total failure...this where the activity assay comes in.
Protein concentration during refolding (in your case by dialysis) is probably the single most important variable.  Try estimating the protein concentration of the guanidine solution after Ni purification by A280 against elution buffer with 20 mM imidazole  (use 1.0 A280 unit = 1mg/mL to convert) and/or by coomassie stained SDS-PAGE (the limit of detection is about 1 ug so load a few dilutions and back calculate from the smallest detectable band being approx 1 ug).  Once you have determined the concentration of the denatured protein adjust it to about 0.1 mg/mL (this is a good starting point but will need some optimization) with 6 M guanidine buffer before dialysis.  You may have to move to dialysis tubing from the dialysis cassette to accomidate the larger volume and use a larger volume of dialysis buffer.
Another variable is temperature.  For some reason 10C has been used a lot.  It's probably not crucial to use 10C, but you should use 4C over RT if you haven't been already.
There is no such thing as one magic refolding buffer for all proteins and (you probably don't want to hear this) some proteins will never refold in vitro.  Several companies sell kits of different buffers that you can sample to find the best one for your protein.  Check the one from Pierce.  Even if you don't want to buy it the web site might be a valuable resource.  The link wont paste correctly, but Its product 89867in the "cell lysis and protein extraction" section. 

louhazosc's picture
Well, I think I figured out

Well, I think I figured out the low affinity. Chalk it up to user error and being a complete idiot. A decimal point was moved over one place too many. Sooooooo....
Everything contained 10X the beta-mercaptoethanol.
Lysis buffer contained 100mM, Ni-NTA equilibration and all washes containing imidazole contained 10mM. Now of course 100mM doesn't go onto the Ni-NTA because this is just the lysis buffer and there is a pellet wash step with 2M Urea before dissolving the insoluble pellet in GuHCl but still, there is probably more than enough to damage the affinity of Ni-NTA once I get to the binding incubation stage.
Buffers re-made and repeatedly called myself a complete dumbass.
However, the advice given on dilluting the samples pre-dialysis is an excellent one and has proven to work well.
Will post results in a couple of days following dialysis.

louhazosc's picture
OK, its getting worse. Now

OK, its getting worse. Now the protein is not binding and coming out in the flow through.
Soooo. Here's some more specifics.
I re-charged the Ni-NTA with 0.5M NaOH pH 7.5 - water-100mM EDTA-water-300mM NiSO4-Water. Equilibrated the resin in 20mM Tris-HCl, 250mM NaCl, 6M GuHCl.
Tried to process it again and it came out in the flowthrough yet again.
It cant be the resin - it just can't be. This is resin that was taken out ofthe bottle, used once, and then regenerated.
This is my last question. I have purified this protein before and it came off at 500mM Imidizole. Then one thing we changed was the lysis buffer. Originally it was 20mM Tris-HCL, 300mM NaCl pH 8.0.
We switched to 50mM HEPES, 250mM NaCl, 1% Tween-20, 1mM Beta-mercaptoethanol.
Finally, upon elution, fractions are taken and buffer exchanged on an amicon centricon with 20mM Tris, 6M Urea to be able to visualize by SDS-PAGE
So my last question there anything in the Lysis buffer with Tween 20 and BME that would be doing something to the integrity of the 6xHIS tag or something going on during the buffer exchange into Urea?
That's basically my last guess.
In order to allievate it if it is an issue. I have already sonicated in this lysis buffer and froze off the sonicate. I am going to try and water wash the pellet 2x.

R Bishop
R Bishop's picture

It seems to me you need a positive control protein that has a HIS-tag.  Im sure there is someone in your lab area/department/email list that has some his-tagged protein sitting around in a freezer.  I would beg borrow steal that protein and quickly run it through your Ni-NTA column in the two buffers you have described here. If that works, then you gotta back up and spike some lysate with the control protein and walk that through. Somewhere in that pathway is the problem.  You are going to have do the painful thing and trouble shoot your way back through the protocol to find it. 
Been there done that.  Best of luck to you mate.

louhazosc's picture
I 100% agree and the control

I 100% agree and the control protein is in the mail from my good ol grad school days.
Down note: Was mailed today and deadline is tomorrow night/wednesday morning. Gonna be some long lab hours.
Thanks for the support

lyndonmungur's picture


I have been working with Ni-NTA resing for a while now with varying results.  These varying results could possibly be due to my protein being a protease, but i have noticed something else. 

My protein has a c terminal His tag and binds very tightly.  As a matter of fact, I too need to elute my protein in as high an imadazole concentrations as 500mM.  It could be closer to 350mM to elute actually, but I use 500mM.

I too was able to bind and elute, but am now finding that the exact same protocol does not work on the same beads, even after recharging.  I am about to try fresh beads, and compare it to the old beads.

In the manufactures instructions, it says that >250mM imidazole is not recommended.  Im thinking that >250mM imidazole is somehow destroying the integrety of the column.  I have noticed a difference in the elution pattern on the chromatogram of the HPLC also.  The conductance pattern has changed dramatically.

Also, are you autoclaving your imidazole.  Ive noticed the change happening after I used autoclaved imidazole on the column (dont know if this is relevent).  Imidazole apparently decomposes after autoclaving.  Maybe this decomposed imidazole is still stuck to my column and dont want to release or maybe damages the comlumn. who knows?  So filter imidazole instead of autoclaving.

Sunil Abraham
Sunil Abraham's picture
 Hv u got the protein and

 Hv u got the protein and whether it reacted in ELISA.

Sunil Abraham