Mammalian protein expression in E. coli - (Feb/24/2009 )

A problem as old as molecular biology itself....

I am trying to express a human protein in bacteria where the only modified amino acids are 3 phosphorylated serines. I’ve gone through the usual hoopla of trying to get it expressed in E. coli (Rosetta/Codon+ cells, varying IPTG, low temperature, etc). Sequencing confirms my insert is correct, but from coomassie gel inspection, I appear to get near zero induction (I should probably do a Western to get a clearer assessment). I’ve heard about custom gene synthesis, and it appears Mr. Gene (https://www.mrgene.com/) would be a good avenue to look into as they optimize the ORF taking into account codon usage in E. coli (though I’m not sure they examine putative mRNA substructure formation like some companies do). It’s only 49c per base pair, so doesn’t seem too cost prohibitive. My only concern is that if this protein is toxic, I could be wasting money.

So I was wondering, for folks who have had similar problems, would you recommend that I first try this over testing protein expression in yeast/insect cells, or the other way round?

Thanks!

If the protein is toxic, you would not see much growth, which you can check by taking ODs after induction.
General questions: What vector are you using? How bad is the codon bias in terms of strings of low-frequency codons? Are there strings of low-freq codons?

For gene synthesis, we have used Geneart and that has worked OK.

If you've tried bacteria a few times, I'd consider moving organism, but you're going to have to re-optimise again (I guess you're well aware of that though...)