Have you checked your plasmid for insert? Sequenced?If you have a recombinant plasmid with GST fusion wy are you getting 25 kDa in WB?

Any how membrane proteins of mammalian origin   are generally difficult to over express in E coli.

Please refer to the following article about expression of membrane proteins.

While working with some membrane proteins, we have observed colonies of different sizes and most of the times small colonies are expression positive. Small colonies are indication of toxicity and anticipate loose expression during scaling up.

Try Autoinduction for protein expression.

Dear Lax

first of all streak your clone on LB plate to be sure that one GST empty plasmid does not  contaminate your work. Membrane proteins are difficult to express in E.coli because  they form inclusion bodies when  are not  toxic. If you do not find  papers on E.coli expression, it is  probable that others have experienced  your problems. Your protein is also  big to be expressed at high level. If what I understand is correct, the GST is at COOH of your protein, and you are able to detect GST by WB.  Then I think your protein should be also expressed, probably it is  in an insoluble form. I do not know your protein nor your project, If your protein is insoluble I suggest to use an His-tag system to use a denaturing method of purification, or move to an eukaryotic system of expression.   Good luck and let me know     

GST tag is a quite a big tag about 26kDa, it is not good idea to fuse to membrane protein. Membrane should be solubilised with 5% Triton X-100 for overnight at 4'C, if you use DDM or DM use at least 1% detergent in buffer in order to see them on gel.

Membrane preparation, solubilisation and purification is described in my paper,s please have a look at these papers

I would also suggest to try with his-tag version.

Best wishes, Thamarai

Article Tripartite efflux pumps: Energy is required for dissociation...

Article Evidence for the Assembly of a Bacterial Tripartite Multidru...

Article Opening of the Outer Membrane Protein Channel in Tripartite ...

Thanks Paduranga,Paola, Lucia, Thamarai for all your suggestions. I was having a close look at them. To clarify some of them: 

1. my construct is good, sequenced and plasmid confirmed by colony PCR.

2. I cant see the article mentioned by Panduranga. Can u once again send that?

3. I can see a WB band ~75kDa when I run the insoluble pellet. That's why I am thinking it to be insoluble. But unfortunately Urea or sarkosy/ Triton 100  solubilization never worked for me to see a band at 74kDa again. 

4. I have another protein with His tag. This is for a double pull down, so I can't change the tag to N terminal or use a His tag for my expression. 

5. I haven't tried auto induction media yet, will try  that for sure.

6. Thanks for the reference papers Thamarai,  I can try using DDM as u suggested.

Will keep you posted. Btw my protein is FBXL2.

Thanks!

Try 5% Triton X-100 for overnight solubilisation at 4'C. DDM is bit expensive. Don't waste at first place, that is why try Triton for longer. You will then see the difference.

Best wishes

Thamarai

It looks like FBXL2 is associated with the membrane via a post-translationally added lipid anchor (http://www.uniprot.org/uniprot/Q9UKC9).  I'm not sure this would be occurring when expressed in E. coli.  You might be seeing your protein in inclusion bodies because of improper folding or it might not be stable. 

Are you expressing overnight at 37C?  If so, I would suggest growing to an OD of ~1.5 before inducing and then express at 18C overnight.  I have found that I get more of an unstable protein by expressing at lower temperatures.  You could also try higher temps like 42C which might stress the E. coli into producing native HSP which could stabilize your protein. 

Another option would be to use a strain that has an extra HSP plasmid or add a chemical chaperone to your culture.  If your protein had a cofactor or a small molecule substrate, that would be the first place to start for chemical chaperones (doesn't look like the case for FBXL2 from the little I have read). 

The final suggestion I would mention is to check the DNA sequence of your gene.  Since it is a human protein, there are probably many codons that are not frequently used in E. coli.  If your bacteria don't have the tRNA to make your protein, it could stall and cause low expression.  You might try a strain with added tRNAs (Rosetta) if you haven't already or you could use a synthetic gene optimized for expression in E. coli. Since you see the correct 75kD band on western I'm not sure this is your problem.  How big is the band? 

I would try temperature first since it is easy to do on a small scale.