Boil the beads with loading dye and run on gel with other fractions.

to be absolutely sure, strip the nickel off the column with EDTA. It will release everything that is bound there, and then try running a gel and see if your protein is there. If it binds too strongly, use cobalt instead of nickel, it binds often in the same manner, but less imidazole is needed to elute the protein. Otherwise, maybe your protein precipitates in the column? Happened to me.

Do you have any method of estimating how much protein you produce in the first place? I mean to be able to separate the expression and purification in order to know which one needs improvement.

I agree with Kamil.

Your low yield can be caused by many reasons, two obvious ones are:

a- your expression system is just not that good

b- your purification is not that good

You should get a feel for how much protein your cells are making, in the crude lysate stage, both soluble and insoluble fractions.

If you know you make a lot of protein, and it is mostly soluble than the culprit must be the purification step.

HTH

I would say that 300 ug/L is terribly low (but of corz it depends on type of proteins we working on). With such amount protein recovery, sometime, we will get a problem on co-purfication of contaminants in the Ni-NTA purifications. I would suggest to solve the problem on protein expression first. Do the expression test by a combination of your plasmid with different hosts. Run this experiments in small scale, for example: 10 mL.

I agree with Kamil on this one (btw Kamil we met in Lund when I was working for a while with Kasia Dymek in biosensors).

As far as I see it you have a problem with either the protein binding too strongly to the column (in which case EDTA should do the trick) or you have a problem during expression. I am assuming you checked the final construct before electroporation. You could run an SDS on the final lysate (after the final centrifugation) and check if you actually have overexpression of your target protein in the supernatant.

Hope my comments were helpful!

Cheers

Hi,

250 mM imidazole seems to me a very high concentration! We use HisTrap-colums from GE and do not exeed 200 mM imidazole concentration for elution. We run gadients from 10 mM to 200 mM.

In that case the protein should be in the first fraction .. and he does not detect it there...

In addition to these excellent suggestions, knowing the level of expression is key, you could also try eluting your protein by decreasing the pH of your buffer instead of using imidazole.

I have had experiences where chelating materials in the media will actually compete off your protein of interest. If you don't see your protein with an EDTA knock off one approach is doing a batch mode purification (mix conditioned media and resin together for an hour or two). Then recover you resin by removing the media supernatant by letting the resin settle, collecting the resin from the bottom of your container, and placing it in a column for purification. One can also add a bit of NiCl2 (2 mM) to the batch to improve binding affected by media chelators. Good luck.

assuming you have a good anti-6His antibody (and you should, as the protein binds so strongly to the column), and you have a detection of >50ng protein, you would need to take let's say 1 ml of your media with secreted protein, concentrate it down to a volume you can run on a gel and then check on a Western Blot, if you have the protein there (and if you maybe have much more).

Of course, one introduces here a problem of whether the WB would work, but that you can check with this small amount you are able to purify: run a first WB with your eluate to estimate the sensitivity. Or if you have another method of detection that is specific to your protein and sensitive enough, like fluorescent tag or sth...

Hello everyone,

Thank you kindly for the replies and suggestions. I'm just using the resin and packing my own columns. @James, I mix the resin with 500mL media + 500 mL of binding buffer (50mM imidazole) and rotate the bottle overnight at 4C. The supernatant is removed the next day, the resin is packed via vacuum flask, then washed (50mM imidazole) and eluted (250mM imidazole) with gravity.

I haven't tried stripping w EDTA, but I have used the recommended stripping buffer (MES) and collected that fraction. I'll run it on a gel to see if the protein is there also. I will also try loading the beads directly. I didn't quantify my expression per se, but after transducing with lentivirus, I screened 12 stable clones for expression with western blot. I'm currently rescreening these clones by accurately counting cells and plating them into 6 well plates to see which ones have the highest levels of expression. I don't need to lyse the cells, the protein is secreted into the media. My antibody is a 6His conjugated with HRP (qiagen) and works very well. How does one quantify expression levels when the protein is excreted into the media? Running a BCA assay will give total protein, but this is useless considering how much albumin is in my media (it's supplemented with 5% FBS). And western blot is not really quantitative for measuring how much protein I get from expression.

I honestly think it is an expression problem, and maybe I didn't screen enough clones. I am using 293T cells for expression but perhaps CHO or NSO would be better? I'm expressing an antibody fragment. Does anyone have experience with either of these lines for expression?

Thanks!

The simplest way is to strip the metal with chelator like EDTA. You will have re-charge your column before next time, but you should remove by this approach everything that is bound on the column due to metal binding.

And you can always compare amount of protein before loading to column and after elution by means of Western.

Boil the beads with loading dye and run on gel with other fractions.

Hi Jason,

First of all sorry for the late reply. To begin with, make sure that the all the buffers that you are using are fresh and are adjusted to the desired pH value. Through my experience, i have observed that immidazole addition some times changes the pH value abrubtly. Secondly, check that the resin is equlibrated correctly with the buffer. Once this is done apply the supernatant and allow it to bind to the resin for sufficient time (at least an hour) depending on your sample volume ( say 40ml from 1L culture). Kindly collect the flow through (after applying the supernatant). Then start the washing step with low concentration of immidazole (20mM) in the wash buffer (600 - 800ml for a litre culture) . You perform this step just to make sure that incase there are other proteins which contain histidine binds to the column will be washed away with 20mM of Immidazole. Collect the wash flow through (just to make sure that your protein is not eluting at this step) and verify it by SDS PAGE. if you think you are loosing the protein at these steps you need to optimize the parameters here before going to elution.

After you verify that your protein is not in flow through or wash then perform the gradient elution ( 0 mM - 500mM) using immidazole . Check the pH of the elution buffer before you start elution. Collect the samples in the fraction collector and verify the presence of the protein using SDS-PAGE.

Pool the protein fractions and then perform the BRADFORD ASSAY to determine the exact protein concentration - try doing triplicates in order to take the mean value ( dilute the unknown protein to 1:10 and 1:100 dilution).

Kindly do follow these steps and let me know if it worked or not.

Best wishes

Hi Jason,

Try using pET28a expression plasmid if you want to optimize at expression level.

best wishes

To check whether there is protein still stuck on the gel, take a little lump of gel and boil it with SDS loading dye and run the supernatant on a SDS-PAGE. It would be better, if possible, to do protein identification by mass spectrometry to make sure the band is your protein. Or you can do an immunoblot with anti-His antibody.

@Kalyana Venneti of course the imidazole changes pH. It's basic substance.

@ Tomas Hluska - thanks for the information. Yeah i do know that.

Hi Kalyana - thanks for the advice and procedures, it has been very helpful. I've used a procedure almost identical to what you suggest. While not all of my buffers are made the day of my experiment, the pH is measured after adding imidazole and I store the buffers at 4C (most are used within a week of being made). I equilibrate my resin (~3mL for a 1L culture) with 50mL of binding/wash buffer (50mM imidazole), then add this to my supernatant directly. The supernatant is diluted 1:2 in binding buffer (500mL supernatant, 500mL of buffer) then resin is added, and rotated overnight at 4C. I allow the resin to settle in the bottle, collect the supernatant, then transfer the resin to a column and pack it with a via vacuum flask. The resin is then washed with 50mL of binding/wash buffer and collect that fraction. Then I elute with 35mL of 250mM imidazole elution buffer. All of these fraction are ran on SDS, see attached gel. I've also done a gradient elution, and ran those fractions on a gel (second image). It's clear from this gel that most of the protein comes off in the first elution fraction with 250mM imidazole, with a slight band after increasing to 500mM. Higher concentrations yielded no more protein.

I dialyze the elution fraction against several changes of PBS overnight, concentrate with an Amicon filter, then run a BCA assay on the concentrate. That is how I obtained the 300ug/L number I mentioned in the first post. So, yes yields are very poor and I think it's an expression problem, not purification. We use a pLenti expression vector with 293T cells, but I believe this is inapproriate for my application (antibody fragments). From what I've read, most people use the GS system from Lonza or a DHFR plasmid, with CHO or NSO cells for antibody expression.

Gradient elution

Hi Jason,

If you want to purify the protein using HIS tag....i advise you to use pET28a expression plasmid and transform it in in BL 21 (DE3) E.coli Strain. This is a better choice and then check the results.

Best wishes

Hi maybe a bit late. But as Sebastian Bartsch said, 250 mM Imidazole should usually be the elution concentration. You should try to run a gradient from 20-250 uM. What you also could try is a step gradient to improve yields.

By the way: What concentation of imidazole does your washing buffer have?

Good luck!

To see if your protein is stucked in the column, I sugest you pass through it some buffer with 8M of urea. If some protein is there, it will be released!

Hi, As Ria did write, run the beads on a gel. Also, you can follow all thye steps by Western Blot (it will also tell you if your protein is cleaved/proteolysed - it happens a lot with mammalian proteins). If you do see that you still do not have a lot of protein, I suggest to change your expresion vector and/or the strains you are using and/or the temperature of overexpression...