Hi Shanice,

You don't actually mention what your problem is. Lack of expression? Degraded protein? Poor binding to your affinity beads/column? Insoluble protein?

OD600 = 0.8 to 1.0 seems a little bit high, I prefer to use OD600 = 0.6  to make sure my cells are still in the exponential growth phase.

20 mM TRIS seems very low. At this concentration the buffer solution won't actually be able to buffer the cell lysate, which can be a lot more acidic than you might think for certain bacterial cultures. I always use 100 mM TRIS.

The pH is a tiny bit low if you're planning to do Ni-NTA (ignore this if this is not the case). For Ni-NTA your buffer should be over pH 7.6 (ideally pH 8 but no higher than pH 9). It will still work at pH 7 to 7.5, but binding is reduced at this lower pH.

The rest of your protocol seems completely fine and therefore you might need to also consider if the problem actually lies with the DNA construct you're using. Sometimes all it takes to make a protein express well, or become soluble and easy to purify is to change to expression vector, or add a solubilisation tag, or change to a different tag, or change the location of the tag from one terminus to the other.

HI Dr Tony, 

Thanks for you reply. The main problem I am facing now is that I am not able to see any difference between the induced and uninduced protein and also between induced empty pET and pET with insert. I have sent my construct for seq and its in the correct frame. 

Hmmm ... if you don't see any difference between uninduced empty pET, induced empty pET, uninduced pET with insert and induced pET with insert, then your protein isn't being expressed.

I assume you have checked the pellet (insoluble fraction) against the supernatant (soluble fraction) of your lysate, which would show if your insert is being expressed as insoluble protein.

Assuming the problem is expression rather than solubility: usually pET28 is a good vector that gives a lot of expression and, since your sequence has been shown to be correct, I think that you probably have a rare codon problem. Rosetta and Rosetta2 already code for many rare codons, but the usage of these codons is still lower than the common E. coli codons and therefore expression of genes with these codons is reduced. If the organism that your insert comes from uses rare codons that might be problematic for E. coli (e.g. Mycobacterium tuberculosis is famous for this problem), then you should consider having your insert codon optimised (this means that the codons of the insert will be changed from the original sequence containing the rare codons, to a sequence containing the common E. coli codons, so that even though the DNA sequence of the insert has changed, it will still code for exactly the same amino acid sequence as the original insert). I have used codon optimised genes from viruses in the past with great success and I'm currently using this method for Thermus aquaticus genes.

If your insert doesn't actually have rare codons, then I would try a few other expression vectors and E. coli expression strains. If this doesn't help either, then I would try homologous genes from some closely and some distantly related organism.

I did check for the rare codon using some software. The rare codon is well covered by pRARE in rosetta. However, I am trying to express a eukaryotic gene (lipase gene from rice) in E. coli. I think this can cause some problems. 

For now, I am trying to optimize the induction condition by trying different growth media, both TB and autoinduction media. I tried to add some glucose to the LB media to stop the lac operon from being turned on. However, I just realised it may cause the LB to become acidic and thus affects the cell growth. Since the TB and autoinduction media contains phosphate bufffer, it may solve the problem.

Yeap. Codon optimisation seems to be the best alternative. But due to time constraint, I dont think I am able to conduct that. I am actually working on the Rosetta and BL21 right now but both neither seems to work.  

Ok ... so my next bit of advice would have been to change expression system. I was going to suggest expression in a eucaryotic expression system that could potentially secrete the enzyme, so I did a quick litereature search and found that this is the right way forward as somebody actually tried this already.

A group from India tried to do exactly the same as you and tested several E. coli strains. They finally managed to get limited expression of rice lipase-II in Rosetta (DE3)pLysS and RIL(DE3)pLysS supplemented with rare codons, but only as insoluble protein in inclusion bodies. Eventually they succeded when they expressed the enzyme in Pichia pastoris X-33, which expresses and, crytically, secretes the expressed protein into the growth medium.

Here's the reference:

Rice (Oryza sativa) lipase: Molecular cloning, functional expression and substrate specificity.

Vijayakumar KR, Gowda LR.

Protein Expr Purif. 2013 Mar;88(1):67-79.

Really appreciate your time and help here. I have read this paper. Just wondering why I cant get expression like them even as inclusion bodies. In fact, some other authors have also successfully cloned and expressed lipase from rice in E. coli. Yes. I am currently working concurrently with Pichia pastoris, struggling to get the cloning done. That is why I still wish to  get the protein express in the E. coli.