Gel filtration purifies by size, not by affinity. So, is MBP close in size to your protein of interest? If so, that's your problem. Use a different purification method- ion exchange, affinity, etc.

Thanks.

But my protein size is ~11 kDa wile size of MBP with His tag is about 43kDa so there is difference in size. But both the proteins are found to elute as a single band.

It sounds like cleavage isn't working.  I assume you are using thrombin or some other protease.  My suggestion is to set up the proteolysis reaction and then remove aliquots at different time points.  Analyze these on a gel by coomasie staining (unless you aren't working with enough material, in which case you could do a western).  You should see the 54 kD band gradually disappear and 43 kD and 11 kD bands appear.  In my experience, the optimal time for cleavage varies a lot between proteins.  One you work it out, then you can scale up for a large prep.  If you don't see any cleavage, the protease or buffer may be bad.  The one caution is that if you let the proteolysis go too long, you will get undesired cuts, so some compromise in the timing may be required.  After the cleavage, I suggest using benzamadine agarose to remove the protease.  You could then reapply your protein to the maltose column to remove MBP and any uncleaved fusion.  The only other explanation for your result is that the cleavage worked, but the two pieces are stuck together noncovalently.  Analyzing on a denaturing gel would help you figure this out.  If that is the case, you might use a higher salt concentration when running the gel filtration column.  Good luck

thanks Michael.

As you mentioned, I have optimized my cleavage conditions for the proteolysis. Denaturing gel shows me two different bands clearly. So I think the second thing, you mentioned could be the reason.

I am using 500 mM NaCl. I will surely try to increase it. Can you suggest me what can be the limit upto which I can increase salt concentration.

Which kind of gel filtration matrix do you use?

Sephacryl S-200 HR from GE healthcare

Are you sure that your Protein runs as monomer? In SDS-PAGE the protein is alsways denatured and of course it runs at 11kDa. But this does not mean, that the native protein behaves same. Did you run some marker proteins, to see how well the column separates the proteins by size?

No. But I did purification of a 17 kDa protein before on the same column and as expected my protein of 11kDa is coming as peak after 17kDa.

I have never gone above 500 mM NaCl, so I can't recommend it.  There is at least the possibility that the proteins are stuck together by a disulfide bond (assuming both parts have a cysteine).  The conditions for loading on a gel (boiling with SDS and a reducing agent) may be more effective at breaking disulfides than the conditions for running a column.  If you want to explore this, you could make samples with and without reducing agent before running the gel.  Otherwise, I think you will have to try different purification methods.  Perhaps put the mixture back on the maltose column (gel filtration should at least get rid of any residual maltose) to immobilize the MBP part and then wash with increasing salt, detergent, reducing agent, etc, Hopefully one treatment will break the association and your protein will come out.  Good luck!

Thanks Michel.

If I use high concentration of DTT would it help to reduce these inter-molecular disulfide bonds? I am using 1mM DTT now in all my buffers.

Does this mean that the single peak - consisting of MBP and your protein - elutes after the position where the 17 kDa protein eluted? Then you should definitely run some standards. Or is this result from a separate experiment with the 11 kDa protein? I agree with Nadine in that the most likely explanation is a multimerization of your protein which migrates similar to the MBP tag. If a considerable amount of undigested protein is still present, the majority of your digested protein could even run as complex with the undigested fusion protein. I would run the MBP-tag alone and the undigested protein alone and compare to the peak of the digested product. Something completely different that happened to me with a similar sized protein: it did not show any absorption at 280 nm.

Ich agree with Michael, there is a slight possibility on disulfide bonds or other unspecific binding. If you use 1 mM (fresh) DTT you should have no disulfide bonds. With 500 mM NaCl I am wondering whether the unspecific bonds can still remain. Denpending on the pIs a native Gel could be an option to find out, what the differences are between native state and denatured state.

Thank you all for your suggestions. Now I am getting single band of my protein by increasing the concentration of NaCl in my buffers.

Thanks once again to all.