1] Check the pI of the protein and make sure you are using the right column [anion / cation].

2] Try to change the buffer accordingly and try to stay two pH units away from pI. if possible reduce the glycerol to 2-5%

3] try using Mono Q instead of Q sepharose.

Cheers

The predicted pI of the protein is 5 and I used a buffer with pH 8.0

that should be good. what is the charge on the protein at pH 8

And, if the second fraction contains your protein predominantly, isn't that a good purification step anyway? Also, if the protein is, for example, His tagged, it will have a very different pI.

Diluting the solution applied to the column may help, and consider using a non-ionic detergent like Tween 20 at about 0.02% (v/v) to help prevent inappropriate hydrophobic interactions among proteins and with the matrix. You might look at the Capto resins which combine hydrophobic and ionic interactions.

Make sure too you've got rid of excess salt too by dialysis or by buffer exchanges on a desalting column.

Superficially, it sounds like you have aggregation of some sort. Addition of detergents can be useful (for example there was a biotechniques paper suggesting a mixture of cholate, Triton X-100 and sarcosyl to fully solubilize proteins. In my experience this can work for things that are interacting with cellular components although for weakly stable proteins the soluble form may actually be denatured. Alternatively screening pH may be a better way. In simple systems the pI of the macromolecule and pI of the resin are all that needs to be considered. However, with proteins there are a great deal of other factors. For example, some of the lipid binding proteins have a very low pI but do not stick to any ion exchange resin unless denatured, while I have worked with a number of proteins, due to their charge distribution bind either to resins that they would not be expected to bind based upon their pI's or can bind both anion and cation exchange resins. What we have found is that getting your protein fully soluble and non-aggregated is the most important step for high quality ion exchange separations and pH, salt, detergent and in some cases reducing agent are the most critical.

Thank you all.

I try to catch up with all your answers:

- The expression level of the protein is quite low (and so far this is the best I can get with all the conditions tried in E. coli using different strains) therefore even if I have a peak with almost 90% of proteins being my protein of interest, I already lost most of it in the first peak giving me very low yield.

- With high salt conditions my protein is not eluted, other proteins come out.

- Since no binding occurs, adding NaCl (even if in small amounts) shouldn't make things worst?

- The lysate is turbid also if the pH of the buffer is 7.6

- The predicted pI takes into account the His tag and the SUMO tag (protein cloned in pET-SUMO vector)

- If Ni-Sepharose column is used for purification, the protein is found again in the flowthrough and very little of it binds.

If I understand your situation correctly, you are loading the protein and getting a major fraction in the flowthrough and less (but pure) fraction during elution. That means your purification is working but most of the protein is going unbound. This could be because of two reasons.

1. You have lot more protein than the column can bind. Check column capacity and load accordingly. Also, determine the salt concentration at which your protein remains bound to the column (say about 300 mM NaCl), and use that condition for equilibration and loading. This will prevent other loos binders in the lysate and allow your protein preferentially.

2. Your column might have some tight binder proteins (from previous runs) that were not removed. This is a common problem with ino-exchange resins) So, do a thorough regeneration of the resin and use it again.

I hope this will help.

Hi. Is there any SDS in the prep? SDS binds like crazy to Q resins and fouls them up. LPS in your prep might be another culprit.

It is not proper to use a turbid supernatant in chromatography. Did you tried centrifugation at least at 20000 xg? If your protein has a pI about 5, at pH 8 (buffer pH) your protein has a negative charge. Do not use a glicerol in your extraction buffer it change the polarity of the solvent. So, if your protein has a negative charge, it must be adsorbed in Q-Sepharose a strong positive charged ion exchanger. You have to avoid using glycerol, get a clear supernatant by ultracentrifigation (23,000 xg), filter through a 0.2 um filter and do chromatography at pH 8 (buffer, 10 mM, pH 8, WITHOUT SALT, NO SALT). You have to measure the quantity of protein you are loading into the column, and also measuring the amount of protein eluting from the column. If with the running buffer you get protein eluting form the column, it correspond with protein no adsorbed to the gel, because oposite charge o because the column is full. You will determine the amount of protein eluted in that condition and then calculate the protein adsorbed to your column. Apply a NaCl gradient (0-2 M, in running buffer) and then see the fractions eluted, and look for your interest protein. Concentrate your fractions (YM10 ultracentrifugation or by liophilization, etc), Dialysate them and make an electrophoretic analysis to look for fraction composition and protein of your interest.

Hector Nolasco

CIBNOR

try 25 Mm tris + 1m NaCl buffer pH 7.2 and conductivity 2 mS/Cm before loading equilibrate the column with this buffer and check the pH and cond of the input and out put....load your sample....and collect flow through...Its always better to do filteration before loading using 0.2 micron filter....and also check the HETP and asymmetry of the column....decrease the flow rate to a minimum while loading....before doing all these u must be aware about the chemical properties your desired protien....and its side chain aminoacids.....check the turbdity of the input sample and output i.e FT....

Why don't you use the affinity tag to capture the protein from lysate? That is much easier to optimize and promises you much higher initial yield.

I'd personally never use Q-sepharose in the capturing step, especially if there is still DNA left. Hydroxyapatite or Phosphocellulose usually perform better.

Have you tried an SP column as well? predicted and actual pI of a natively folded protein could be very different.

I will try tomorrow again with the Q-Sepharose column using only buffer (20 mM Tris/Cl pH 8) without any glycerol and I'll let you know

Ion exchange is a little tricky. First your sample has to have low salt ie less than 50mM NaCl even lower if you can manage it. Second the column should be equilibrated in low salt conditions (ie 0mM NaCl). You apply your sample slowly and use 0mM NaCl for a column volume or more to wash anything that does not stick. If your sample comes out under those conditions then Q-sepharose is the wrong media use cation exchange resin like CM or SP. Also get rid of the glycerol it may interfere with binding could unfold your protein and increases the botany density or you lysate making it more difficult to pellet your lysate.

I agree with Bryan but I add...probably your protein exist only in aggragate form. I hope I'm wrong.

I agree with Daniele, your proteins seems to be in aggregated form as you have mentioned that the solution is turbid. So the isoelectric point of your protein is not the same as the pI of the actual protein. In my view, there is no harm in trying hydrophobic column.

And I think Rob is right, if you are getting mostly your protein in the 2nd fraction then you have the protein. To purify further, you can go for gel filtration chromatography (after concentrating the flow through on centricon), collect the peaks and analyse on Western blot with specific antibody.

All the Best

If your protein is aggregated, it may not bind to the column. Try to clarify your lysate by ultracentrifugation (e.g. 100,000 x g) and see if your protein stays in the supernatant as a test for aggregation The turbidity of your sample may be due to aggregation of your protein. You may need to put some salt in your lysis buffer to prevent aggregation, although this comes with the risk of competing with your protein for binding to the column. Some non-ionic detergent might be helpful. If you are using an overexpression system to produce your protein, try reducing the level of expression to improve the folding of the protein.

I agree with Nitin, you need to check your protein to be sure it's not aggregated form. Because if your protein is aggregated form it seem difficult to manage in down stream process. Even though your protein was eluted in 2 fractions of Q-sepharose but for me it seem good because almost of 2 fractions contain of your protein. but you need to work more by done the second purification. Maybe you can use gel filtration column for second purification, and dont forget to concentrate your protein from the first purification before loading in the second colunm. ^_^

After I sonicate I always have a ultracentrifugation step (JA25.50 rotor for 60 minutes at 20K RPM) and I also filter the supernatant and it is still turbid.

Yesterday I purified with 20 mM Tris/Cl pH 8 and still my protein is in the flowthrough.

Should I try changing the buffer to phosphate buffer?

Try putting that flowthrough onto a fresh Q column. If you have high LPS conc. in your starting prep, it will bind tightly to the Q resin and compete for the protein binding.

Do you use DNase during cell lysate preparation and how much? I am wondering the turbidity of your lysate might be due to high concentration of DNA, which subsequently prevents your protein to bind to the Q column.

Turbity of your samle indicate that your protein is surely aggregated. The first step would be to bring it to a soluble form (you are using glycerol for the same purpose which acts as a chemical chaperone). The protein can be de-aggregated by using a bit higher concentration of glycerol or some other chemical chaperone. after solubilization the protein sould bind to the Q-column as it is a strong anion exchanger. Also do not forget to use glycerol in the euilibration buffer or binding buffer (in case it solubilizes your sample).

A soluble protein solution will look like "GOLDEN" in color whereas an insoluble one appears as "CLOUDY". get rid of that cloudiness first...

good luck

I use DNase in the concentration of 0.2 mg/ml. I add it after sonication.

Ciao, don't waste more time, try new fermentation parameters (T°, medium indiction time and so on) in order to reduce the aggregation state from the beginning. let me know if you need more information aboit it.

Ciao Daniele, this is so far the best expression system that gives me soluble protein. I will switch to Pichia pastoris soon but the point is that anyway my protein in not unsoluble since after sonication and ultracentrifugation I can find it in the supernatant (confirmed by SDS-PAGE). Or am I wrong?

Probably I missing some information, you protein comes fro eukariotic system?

yes, it's from a fungus but I'm expressing in in E. coli BL21 (DE3). The gene was codon optimized for E. coli expression and the expression system is pET SUMO from Invitrogen (http://products.invitrogen.com/ivgn/product/K30001)

Some points to consider:

-if you suspect aggregation use Urea from 3 to 5 molar during binding and in equilibration buffer.

-as DNA is a serious competitor for anionic exchange, consider a protocol for its elimination or increase the quantity of Q-sepharose used.

-after equilibration of the column, check ph and conductivity of the flowthrough of the column

-Check pH and conductivity of your sample as well.

-if your protein is calcium or other metal binding protein, then the theoretical pI do not says so much. Use EDTA around 5 mM in your buffers.

my PI suggested to stick to the same buffer and add a reducing agent and run the flowthrough on a SP-sepharose column

I don´t think that turbidity is due to DNA, maybe protein aggregation, use a chaotropic (urea, GuHCl) or glycerol as it is suggested above or ultracentrifuge the sample. Maybe the aggregation is not due to your target protein but if it is your protein you could have it nearly pure and you can subsequently solubilized it with 6-8 M urea and try to refold. If you want to eliminate DNA before performing Q chromatography, you can use PEI, a cationic polimer wich is able to bind DNA very efficiently. You should incubate with the polimer and centrifuge afterwards to eliminate PEI-DNA but....be careful because if your protein binds DNA probably it will go with PEI also and sometimes it is difficult to recover the target protein.

if your protein is as aggregates it won´t be able to bind to the resin unless you use urea. You did not mention if it is expressed very much in E. coli, maybe you have IB, there are several good protocolos in order to solubilize them.

Good luck!

i am sorry if i have not got your query correctly.

you said that you do get a fraction, albeit in the flowthrough which is your pure protein and i believe that is what you want ?

also predicted pI may not help as the actual pI is dependent only on the surface amino acids.

the suggestion on the turbidity due to DNA is valid but i would try to bypass it with adding benzoase during the lysis protocol. i have tried to use PEI but the results have been inconsistent.

if i were you i would proceed with the second flowthrough fraction

good luck!

Hi there, I think this is an easy fix because I've run into the same problem. You need to get rid of the glycerol, it interacts with the column and prevents protein binding. I drove myself crazy and decided to eliminate one by one the ingredients until I found that getting rid of glycerol allowed binding. Also, your protein may be precipitating from no salt concentration, add 50 to 150 mM NaCl and it should still bind to the column.