Hi Shoaib,

you already performed some decent expression tests. I would suggest two more options. First, try an E. coli strain providing a less reducing milieu in the cytoplasm: some Origami (DE3) or ShuffleT7: Expression of single-chain Fv fragments in E. coli cytoplasm DOI: 10.1007/978-1-60327-302-2_17

Second, as you already mentioned, go for a fusion protein: Maltose-Binding-Protein or Alkaline phosphatase: A simple vector system to improve performance and utilization of recombinant antibodies DOI: 10.1186/1472-6750-6-46

In this context also very nice is a Thioredoxin fusion in combination with one of the above mentioned strains: Overproduction of anti-Tn antibody MLS128 single-chain Fv fragment in Escherichia coli cytoplasm using a novel pCold-PDI vector DOI:

10.1016/j.pep.2011.12.010

Expressive regards

Thomas

Hello! When I have problems like this I worry about the cloning in the plasmid and re-sequence. My first thought is to isolate a clonal colony by streaking for purity, after transformation of the plasmid stock (into DH5 alpha), followed by simultaneously making a glycerol stock of the clone and doing a plasmid prep. You can use BL21 for a plasmid prep if you have to, but often labs will have a DH5 alpha (Top10 or other) workhorse strain for cloning and your plasmids are safer in it.  If you are sure of the cloning, I would only do three isolated colonies, if not sure five. I would send for sequencing and check my cassette for insertions or deletions. The kozak sequence or secretion sequence may still be attached. I would remove them with 'quickchange' PCR primers. An ScFv should express in BL21. As a rule I sequence any plasmid given to me by someone else. I also make glycerol stocks because the plasmid is more stable in the bug. For a protein I want to do several protein preps from I will make a BL21 glycerol stock from a verified plasmid sequence.

     Another thing to consider is that the BL21 is not 'curing' your pET28a plasmid. Please be sure that you have at least 50ug/ml of Kanamycin in your expression media. I think you can also get away with 60ug/ml on the plates. However, the most likely candidate is that there is something wrong with the sequence. You can also check the codon usage,after sequencing, with The Graphical Codon Usage Analyzer. http://gcua.schoedl.de/ This is a good E. coli codon usage frequency table, http://www.faculty.ucr.edu/~mmaduro/codonusage/codontable.htm Good luck!

http://www.kazusa.or.jp/codon/

Hi Shoaib,

I am the primary author of the the paper that Thomas mentioned about. I had tremendous expression trials when I first worked with scFvs. If you use the cold expression (18oC) system using pColdI vector,  fusion tag of Protein Disulfide Isomerase and induce with 0.5mM IPTG at lower temperature 18oC, you will get almost 95% soluble expression of scFvs. Also very important is the use of Origami B(DE3) bacterial strain which has oxidized cytoplasm for expression.  Please note that Maltose Binding Protein (MBP) or Thioredoxin (TRX) fusion tags did not work in my hand. The expressed scFvs will have disulfide bonds formed with correct cross linking between the cysteine residues in each scFv domains using this system.

You can read more about it in: 

Protein Expr Purif. 2012 Mar;82(1):197-204. doi: 10.1016/j.pep.2011.12.010

Goodluck,

Ganesh

Hi Thomas, 

Thank you for your suggestions.

Hi Ganesh,

Thanks for suggestions. I will give a try to your vector, tag, strain system and I hope it will work.

Hi Matthew,

Thanks for your valuable suggestions.

To me it sounds strange that you cannot detect the scFv, even by WB. As mentioned above; are your sure that your DNA is as it should be? Sometimes the problems can be solved by transforming fresh plasmid into fresh BL21. A simple trick that you can try is to plate your transformed cells on selective agar plate. Next day, use all of the colonies to inoculate an 80 mL TB culture already holding IPTG. Sometimes this works for proteins that are toxic for the cells, and it is very easy to do.

So long story short, I think that there is a simple solution to your problem, because it does make sense that you cannot detect the protein by WB analysis.