Toxicity of target proteins in pET vectors can manifest itself without DE3. Some people suggest E. coli polymerases causes low level of expression. The observed mutation rate is thousands of fold higher than what the textbooks say. Its easy to tell toxicity from other causes of heterogeneity. Less than expected numbers of colonies from transformation. The growth rates of cultures from different colonies are quite different. Most mutations are single nucleotide change (and single amino acid change) that reduces certain functionalities of target proteins. However the mutations can be outside those functional domains. In my experience with a few toxic proteins, the mutations are evenly distributed, no over representation of any mutation. Isolated wild type clones generate mutants after transformation.
Chun From: CCP4 bulletin board [mailto:CCP4BB@JISCMAIL.AC.UK] On Behalf Of Rubén Sánchez Eugenia Sent: Wednesday, January 25, 2012 8:52 AM To: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] Expression of Viral proteins for crystallography Dear Gregory and Darren, Thank you for your answers. They have been very useful. 2012/1/24 Gregory Verdon <gregory...@gmail.com> it possible that the protein is toxic (even when slightly expressed from your possibly leaky pET vector), so that e.coli select for mutations that kill expression of your recombinant gene ... Gregory, the protein may be toxic, but I do not think that could be the problem. BSJ cells are not meant to be for expression, but for cloning, so they are not DE3 lysogenic. That is, they do not have the T7 RNApol necessary for the expression of the proteins inserted into pET plasmids. In any case, thank you very much for you help. 2012/1/24 Darren Hart <h...@embl.fr>: > I think the explanation is this: > The source is natural viral RNA which is a mixture of naturally mutated > sequences (e.g. flu forms such a quasispecies) > See: > http://www.virology.ws/2009/05/11/the-quasispecies-concept/ > > The pooled RNA has an average sequence that you see when you sequence the > pooled cDNA (individual mutations are hidden by the averaging effect of > having many sequences present). > > But when you clonally separate DNA molecules by transformation (1 plasmid > enters 1 cell to yield 1 colony), you see each individual molecule > represented 100% in the sequencing chromatogram from the plasmid DNA that > you have isolated from colonies. > > This is effect is commonly observed when sequencing influenza virus isolates > from patients. It will have nothing to do with the E. coli strain. You can > avoid it completely by using gene synthesis. > > Darren Darren, thanks to your information about the quasi-species, now we are convinced that this may be the problem. We have been thinking about this and we have conclude that the average sequence may not even exist and if it was, it might not be the active one. That is to say, maybe we can not synthesize our gene because we don't know whether the mutations are needed or in any case what mutations would be needed. Do you agree? And if I am right, is it the only solution to try different mutants to get the active one? Anyway, I think I would not need to synthesize the gene because I can select one monoclonal colony and delete the mutation by site-directed PCR mutagenesis. Best regards,
