Maybe it is a complex or if the peak is around the void volume of the column it could just be aggregate?

Hi. This is not unusual. Gel filtration usually does not fractionate proteins into very sharp discrete peaks, because protein shapes, lipids, detergents, etc. also influences how they run. The single peak you are seeing is probably a convolution of peaks from proteins, protein complexes and other assorted stuff.

Ion exchange could be a better next step.

What kind of mobile phase (buffer) did you use? It helps if you add the detergent into the mobile phase and sample (NP-40, triton x-100, cymal,....) It looks like you have aggregates, so you don´t get the separation on gel filtration, which is often with membrane proteins.

Dear Sean,

The column volume is 24mL, and the peak is at 12.48mL, actually there is also a small peak at around 8.88mL, which should be the aggregation peak.

The molecular weight of the protein is 58kDa, and I used 0.025%DDM as detergent, based on former experience, 12.48mL is a reasonable peak position.

Higher molecular weight bands could be aggregates and lower molecular weight bands degradation products, but if none light up with your anti-His C-terminal antibody, they may be contaminants. Was the single peak in the included volume or in the void? What is the estimated molecular weight of the major peak; many membrane transporters are dimeric. Was the Ni column new? S-200? Perhaps it was contaminated from previous runs. DDM should be well above cmc, try 0.1%. Also, try different detergents for solubilization and check your protein for solubility and mono-dispersity.

Maybe in pooling fractions you are collecting some fractions from the pick that came out in 8.88 ml. Did you check all fractions on SDS PAGE gel?

when you ran your gel, how did you prepare your sample? For a membrane protein I would heat for a short time at 70C, rather than 100C. You might be degrading your protein before loading on the gel. How pure was your sample before you loaded it on the GF column?

Since it is a membrane protein, high chances of protein getting degraded. This might be happening after gel filtration cinematography.

Thanks for the suggesitons!

I have tried to solubilize the protein in 6 different detergents: DM, UM, DDM, Cymal6, OG and FOS12. And DDM was relatively the "best" (I still lost half protein in this process), I have also screened the concentration of both membrane and detergents and solubilization time, but nothing helped.

The Ni-NTA agarose was new. As for chromotography, I used 20mM HEPES pH7.5, 100mM NaCl and 0.025% DDM as mobile phase. The column volume is 24mL, and the peak is at 12.48mL, actually there is also a small peak at around 8.88mL, which should be the aggregation peak. The molecular weight of the protein is 58kDa, and I used 0.025%DDM as detergent, based on former experience, 12.48mL is a reasonable peak position.

For the last membrane protein I worked with, Ni-NTA and Gel Filtration two steps purification was OK, but for this one, it looks like I have to try ion exchange chromotography.

Just one "small" remark. When integral membrane proteins in detergent micelles are run on SEC the size of the detergent micelles must also be taken into account. To my best knowledge the size of an "empty" DDM micelle is around 20kDa. Thus a single trans-membrane protein with 8 TM segments and 58 kDa size would show up on the SEC as a globular protein with molecular mass of around 90 kDa.

0.025% DDM (=dodecylmaltoside?) might be too low. I have never used DDM at concentration lower than 0.5%.

Sorry; lower than 0.05%.

You should consider two possibilities

(1) the MP is being cleaved during the prep but is kept as one due to the DMDM micelle surounding it.

(2) The MP is not cleaved but when in SDS at the concentration you are using it solubilises variably. In otherewords the helices separate out into the SDS micelle with anounber of different associations with SDS so you get a range of discrete MP+SDS complex.

I has seen (2) sometimes when the MP is very hydrophobic.

If you do a MS molecular determination you may spot if it has been degraded.

I had similar experience. I was able to remove all the contaminants by using more stringent Ni-NTA washing buffer (50mM immidazole in my case) and get pure protein. Also as other people suggested, you might be getting a mixture of uniform-sized micelles that contain many contaminating membrane proteins. What do you need the purified protein for? You make consider purify the protein under denaturing condition and then renature it on column. This way the purity is guaranteed.

More stringent washing with imidazole (50mM) before elution, sbbsequent elution, then dialysis; an ion excahnge before the gel filtration might be helpful to get your 58Kda band. That will also be helpful for getting rid of lipids and other associate proteins which might appear as background, later you can try the gel filtration. Hope this helps.

With membrane proteins, because of solubilization and using detergents, gel filtration is not typically recommended. With detergents, micelles are formed and as Steingrimur as aptly pointed out, there could be many proteins co eluting with your protein, ie because it is either with your protein in a micelle, or just co elutes with it because of hydrophobic/hydrophilic interactions. If the detergent concentration is below the CMC, the proteins still don't typically migrate at the right MW. Also, other proteins co migrate because in your system, they just have the same MW. So during the purification, I would recommend doing the Ni NTA first, to remove almost all the contaminants, and then using ion exchange. With membrane proteins there are limited options. But ion exchange usually works nicely. With the ion exchange, you will likely have to do a quick buffer exchange of course.

Marcia

I am also working with membrane proteins and had multiple bands on the SEC column and on the SDS. I have changed from LB-Media to minimal media expression like the paper "A method for efficient isotopic labeling of recombinant proteins", Marley 2001. Also without isotopic enriched nutrients. So I got a cleaner sample after His-Tag purification and it reduced the oligomerisation of my protein.

I have no experience with E.coli Top10 cells. My favorite is E.coli Tuner (DE3) at the moment. Also quite good regarding basal level expression. E.coli cell disruption using microfluidizer and sonifier only was also not sufficient enough. I could extract lipids in my final, concentrated sample. To prevent the oligomerisation caused of possible E.coli lipids, I tried lysozyme before disruption. It helped.

For splitting oligos/aggregates - what ever - freezing and melting or refolding and renaturation is also a possibility - if your proteins goes with this rounds. I recommend refolding in terms of dropwise dilution in a buffer with detergent, similar to "NMR Structural Investigation of the Mitochondrial Outer Membrane Protein VDAC and Its Interaction with Antiapoptotic Bcl-xL" Malia 2007. In the line of detergents there are LDAO, Brij 35 and the good old CHAPS which I can recommend. Brij 35 has a very low CMC (0.011% w/v). I am using 0.005%. It has splitted my oligo to a dimer (Dimer is good in my case). The disadvantage is you can´t dialyse it.

One remark to CMC: Some proteins make oligomers in the presence detergent despite a concentration lower than the CMC. It depends on the detergent. It was shown for Bcl-2 family members like Bax.

Affinity steps work well too if ion exchange doesn't work or you need additional purification. When I purified a membrane bound enzyme that was not tagged, I used an affinity step based on an inhibitor of the enzyme. So depending on what your 7 TM binds to, you can either add a agonist/ antagonist that can or is either attached to a column, or has a biotin tag. With a biotin tag, there are ways to detach the ligand and your 7 TM from the column. Or just compete it off with another agonist or antagonist.

Dear Rhine Nieh, your results make sens to me the multiple band is likely due to either different oligomerization state of your protein, (for MW of empty micelle is depends from protein to protein ) I would not be surprised if from MALS analysis you got a trimer of your protein but could be a dimer as well, the other option is that you might have a binding partner but I would bet on that however some or your overexpressed protein might have lost its tag. You can also check the oligomerization of your protein by the doing a co-expression of the same protein with a different tag (7MX-8His and 7MX-strep co-expressed for example) and you purify via his-tag affinity and identify the purified protein using the second tag.

i have a little experience on same problem.there is possiblity dat many proteins exist in the sample may be having same size of MW consequently gel filtration profile has to show one peak only where as in case of SDS page protein may get degraded as aresult they will show multiple bands or u can also opt for 2d gel.

1st

Your protein might have come off the column in the void volume. The void volume would contain all proteins and/or aggregates that are larger than a certain size (in your case larger than around 2000kDa). When boiling in SDS sample buffer, those components stop interacting, hence - arange of bands in the SDS gel.

2nd

Your protein is unstable and is being degraded. During size exclusion chrom. however, the degradation product still stick together and only seperate in many bands after boiling in SDS sample buffer

you should check how clean your sample is before loading onto Size excl chrom column. If it is rather dirty and contains a variety of protein bands (assess via SDS PAGE), dont bother doing Size exclusion, as this only serves as "polishing step" after a "neat" Nickel IMAC.

Plan subsequent purification steps after NickelIMAC according to how clean your NickelIMAC prep was. You could do Ion exchange before trying Size excl chrom.

Using gel filtration of a pure protein give you an idea about the molecular weight of your protein ,if you run, through the same column using the gel a collection of standard proteins, then if you run an SDS electrophoresis of the only band that you elute from the column ,then this will give an idea if your protein is a monomeric one or oligomeric one depending on how many bands you can get on the electrophoresis gel & their molecular weights. I think in your case you need

1- purify your protein .

2- determin its molecular weight using gel filtration.

3- do SDS electrophoresis using your purified band after boiling it for 2-5 minutes in presence of urea ,mercaptoethanol and SDS before applying the sample on the gel to be sure you break any hydrogen bonding & S-S linkages present within the structure so there is no interactions between the subunits if your protein is of oligomeric type.

Good Luck