One possible reason for low expression is the presence in the gene of rare (from the perspective of E. coli) codons, for which tRNA abundance may be a limiting factor. You can either have the gene synthesized with optimized codons, or use an E. coli expression strain such as Rosetta that has the genes for tRNAs for rare codons.

You can't usually get high expression of a transmembrane protein in E. coli, especially when it's from a different kingdom.  "A small amount" is actually really good (as opposed to nothing).  

E. coli is best for expressing simple, soluble proteins from other bacteria.  You can always get lucky, and then you get publication bias of course.  But the more different your sequence is from simple, soluble, bacterial, the less likely you are to get lucky. 

If I was trying to express a transmembrane receptor, I'd do it in cell cultures of the closest species that has well-characterized, easily available cell cultures.  For example, human receptors in human cells. 

The usual suspects for low expression are protein toxicity and codon bias. And I agree with John that it is more easily to express a transmembrane receptor in cell cultures of the closest species than in E.coli. The following link may help you.

The problem of protein toxicity may arise when the bacterial protein performs an unnecessary and detrimental function in the host cell. Most expression problems are also the result of protein toxicity. With optimization of expression vectors and host cells, we estimate that 80% protein yield problems are the caused by protein toxicity.

The problem of codon bias is also a common reason of protein expression problem. By codon optimization, we can improve protein expression level in living organism by increasing translational efficiency of target.

https://www.researchgate.net/post/A_pET28a_bearing_an_insert_is_transformed_into_E_coli_BL21DE3_But_the_protein_is_not_expressing_with_conc_of_IPTG_and_different_temperature

Hi there,

Over expression of membrane protein is problematic also because it pertubs membrane composition. To overcome this issue some E. coli variants named C41 and C43 less susceptible to the effect of massive insertion of protein into membrane have been isolated. In this background yield of protein production should be improved. See the following paper:

http://www.ncbi.nlm.nih.gov/pubmed/8757792

 All of the above discussions are very relevant. I would also like to add my thoughts

1. If you pursue expression in E.coli, obtain a suitable strain for the job. As discussed above there a a multitude of strains available: disulphide bond tolerant, codon bias, tandem repeats etc. 

2. If your protein is toxic upon expression then you will have to regulate expression much tighter. Again there are strains available for this, see BL21 strains/pET plasmid combinations. Increase bioamass without induction, then induce at a later stage. This way if protein is toxic and inhibits cell growth you already have the biomass present

Note if using BL21 strains always transform with fresh plasmid, do not use glycerol stocks.

3. You could use a heat shock prior to induction i.e hold culture on ice. This induces chaperone protein expression which may aid in protein folding

4. Use a different promoter (weaker promoters maybe better). 

5. Play around with the use of different media

6. Play around with induction length. 1hr + 

Thank you all for the advices, I currently expressing using C41 e.coli strains and searching for Rosetta and also BL21-CodonPlus e.coli strains(just in case), hopefully it gives me some good expression amount for purification, update soon.