Protocol Online logo
Top : New Forum Archives (2009-): : Molecular Biology

sacB Counter-Selection to generate double cross-over deletion mutants - (Mar/20/2013 )

Hi all,

I am making deletion mutants in Pseudomonas aeruginosa by allelic exchange. Basically, I construct a deletion cassette with gentamycin resistance (~700 bp upstream portion of gene----Gentamycin (Gm) Resistance----~700 bp downstream portion of gene) and I clone this into a suicide plasmid with Carbenicillin (Cb) resistance and sacB (Counter-selective measure that inhibits growth in the presence of sucrose in gram-ve bacteria).

I electroporate the plasmid into P. aeruginosa, and select for transformants on LB Gm 30 ug/ml plates. I then patch isolates onto LB Gm 30 ug/ml AND LB Cb 200 ug/ml to differ from single or double crossover. I assume Gm resistant and Cb resistant isolates are Merodiploid, while Gm resistant and Cb sensitive isolates are double cross over events (gene KO).

In my case, I have only seen merodiploid isolates (Gm and Cb resistant), telling me the plasmid is still integrated and I have a WT copy AND a deletion copy still in the chromosome.

To resolve the merodiploid isolates by sacB counter-selection, I have tried growing merodiploids in LB Gm 30 ug/ml and 5% Sucrose, then dilution plating on the same media. I get isolates, however they remain both Gm AND Cb resistant.

Is the counter-selection not tight enough? Also, I believe there are two copies of the gene, so I need to tighten up selection to ensure integration/ recombination at both sights. Any suggestions?

I plan to run a control LB Gm + 5% sucrose plate to determine the level of inhibition caused by sucrose. I will run a colony PCR once this issue gets resolved and will update.

Please, any help is welcomed!


sometimes the counter selection is just not stringent enough...we have the same problem. check out this paper, "Flagellin A is essential for the virulence of Vibrio anguillarum.", they select the second crossover by looking at the morphology of the colony but I never figure out exactly how they did that.

-Lynn Shay-

Counterselection is difficult at best. The sacB counterselection works rather poorly, in particular. The major difficulty is that you are working against evolution -- living systems try very hard to survive, and it is easy to knock out a counterselectable marker (or its effects) in many other ways than they way you would like. A good counterselection combines an selection and counterselection in the same marker, which makes the galK knockout good. Selection on galactose medium requires the marker (native galK) to work, while selection on deoxygalactose is lethal (again requiring the galK gene to function).


Hi all,

Thank you for the information! I haven't noticed a difference in colony morphology; thank's for the paper. The galK marker sounds like a great system, but I am already invested in using sacB and have some useful updates.

A colony PCR using CDS primers (amplify gene from start to stop codon) revealed that my isolates have two sized amplicons; once of which is the WT sized band. To determine if these isolates resulted from a single cross-over (they contain plasmid backbone with a KO copy and WT copy) or a double cross-over (they contain a clean KO but there is a second WT copy of the gene) I amplified sacB. The isolates do not contain a sacB gene as per colony PCR. This indicates the presence of a second copy of the gene, which I know was probable from the literature.

I am thinking to repeat the process once I remove the Gentamycin marker from the KO copy. I will remake a KO cassette that targets the internal region of the gene (that is now only homologous to the remaining WT copy since this region has been deleted fom the KO copy). Does anyone else have any ideas?

Thank's again.


I've actually been trying to get some deletion mutants in PA14 using the same method you are using. I generated my recombination cassette using the method described by Choi et al. with 500-1000 bp homology regions flanked by a Gm cassette. I do a tripartite mating with the recipient strain, donor strain containing the homologous recombination plasmid (Gm, Cb, sacB), and pRK2013. I first select for Gm resistant mutants  on LB+30 ug/ml Gent, and then I streak all Gm resistant colonies for single colonies on 5-10% sucrose + Gm (30 ug/mL). You will notice that some colonies, when restreaked, do not give single colonies and instead grow very rapidly and diffusely at 30C. These are, for whatever reason, still merodiploids and are Cb and sucrose resistant. However, there will be some that grow noticeably different and more slowly giving rise to single colonies. These are most likely your deletion mutants. Patch on Cb as well as Sucrose+Gm to verify their sensitivity. Then, do colony PCR to verify deletion of the allele. Hope this helps.


Dear Gandhi,


I have been working on my PhD project for a year and a half now on Pseudomonas aeruginosa PAO1 using this exact same technique. The final step seems to have been giving every one I know a lot of trouble. Not a single person has managed to get their deletion using this technique just because of this last screening step. A lab mate of mine gave up on his project because of this last step and started a new one. 


When screening, I get the perfect phenotype, but then colony PCR proves me otherwise, giving me only wild type gene. Not even the GM resistance gene amplifies even though the cells are growing like crazy on 30ug GM. The same lab mate suggested that it might be the pH of the medium and suggested MH. I got my transfomants growing on MH 30ug/ml GM this morning and now I am going to screen on LB 30GM/5% suc and LB Cr. Tomorrow I will be doing colony PCR, wish me all the luck. 


If any one knows any one who managed to get their deletion using this technique, please let us know. Other than Choi of course ;) 


will update





Ok, not even MH worked, same colony PCR results as on LB :( this is just not working. 


Ok I am going to post this incase someone still needs it. I got mine to work after a lot of advice from a fantastic professor that I got in contact with. Select on Carbanicillin 200 ug/ml. Forget about the GM for now. For PAO1, Carbanicillin works really well. You should get a few colonies, streak them on 8% sucrose, yes 8%! Then run colony PCR get the ones the give you the mutant gene and then grow them on Pseudomonas isolation media (PIM) with 30ug/ml GM and 5% sucrose. Take a single colony, I repeat, a single colony, grow it in rich media broth with 15% GM. The next day, do a serial dilution and take the 10 to the minus 8 and 10 to the minus 9 dilution and plate them on rich media. Run multiple colony PCRs on various isolated colonies to make sure they are what you want them to be an there you go. Make sure to put plenty in 20% glycerol then store them t -80. 


Hi Abdool,

                  At what temperature did you grow the rich media broth with 15% GM at 30 or 37 degree? Are your mutants that you got through this method Carbanicillin sensitive?