My research group is working on integrating ampicillin resistance cassette into K-12 (MG1655) strain of E.coli but without success. It is interesting that ampicillin resistance gene is generally only used on plasmids. Does anyone know why this is?
plasmids are multicopy which authorizes resistance to high ampicillin concentrations. May be a single chromosomal copy of the resistance gene is not sufficient with with your ampicillin working concentration?
I would agree with Dr. Levesque. As a preliminary experiment you can dilute your ampicillin for tens of folds and see if there is any significant resistance.
Actually I have to disagree with Herve Levesque , you can integrate the amp gene as a single integrant in E. coli and it will confer resistance perfectly well. Many natural plasmids carry the amp gene on a transposon (Tn1 and Tn3 as examples) and there is extensive literature using these transposons in E. coli where they integrate in the chromosome. This was first observed in the late 1960s to early 1970s and was used extensively as a genetic tool for the subsequent 20 years or so. I have done it myself and in my lab numerous times. The bias you are seeing in the literature is merely a reflection of the types of experiments people are doing, there are LOTS of people cloning genes in E. coli but many many fewer doing genetics and such in E. coli.
Dr Benedik is right, chromosomal mutations confer resistance to antibiotics and mobile genetic elements like transposons support such horizontal transfer. The question was rather the degre of resistance of chromosomal versus (multicopy) plasmid resistance genes : to my knowledge, the second generally allows to resist higher antibiotic concentrations, but I may be wrong!
Herve Levesque is correct that you can get higher resistance to some antibiotics when present at higher copy number. But not to all. I'm not sure about with ampicillin but single copy confers very good resistance. Chromosomal amp mutations are pretty low level (10ug/ml or so) but single copy transposon certainly confers resistance to 100ug/ml or more, depending upon strain background.
But to answer Helga Kristindottir more directly, using the typical range of 50ug/ml of amp should be just fine for selecting a single copy of the bla gene in MG1655. So if you are unable to get it, the problem likely lies elsewhere in the protocol.
If you care to share more experimental details someone here might be able to comment or help troubleshoot.