Markerless mutants using suicide vector

[dhkwak1]

Hello,

I'm wanting to generate markerless deletion mutants in my bacteria (A. baumannii) and am using a suicide vector (gentamicin resistance, SacB gene). My vector contains ~1.2kb of homologous sequences that flank both sides of my gene (~2.4kb homologous regions total). After selection on sucrose, I've found that 100% of my clones so far are WT and not the deleted mutant as desired. I'm wondering if anyone would have any suggestions on troubleshooting these types of issues? One specific concern I had was the GC content of my 5' vs. 3' flanking region and how that could affect the frequency of crossing-out to yield mutant vs. WT allele. Thanks a lot for your time and let me know if anything I've said is unclear.

[JackBean]

Are you sure that the bacteria is capable of homology recombination? Is your vector linearized? What about usage of two linear pieces which are homologous to each other?

[dhkwak1]

Thanks for your response, JackBean.

Markerless deletion of alleles in this bacteria have been reported (a few 10's of times in the literature) although it is reported that illegitimate recombinations are more frequent in this bacteria than other gram-negative bacteria. We had been able to generate a single mutant using this method before, but we're having trouble deleting a gene within the same region (operon), and the issue (and difference between the two) are not so obvious for us.

Our vector is a circular plasmid. Can you elaborate on what you mean by two linear pieces that are homologous to each other? Do you mean using a linear DNA to KO the gene of interest? If so, we had actually successfully applied this method to our bacteria.. once (it couldn't be reproduced!). However, our linear KO DNA also contain an antibiotic cassette that we suspect may introduce some polar effect on downstream genes, so we are trying to avoid that route. If we try to, for example, use linear DNA without some selectable marker, we'll just have a difficult time screening for our mutant.

One interesting detail: the GC content of the upstream and downstream flanking regions used to recombine with the flanking regions of our gene of interest for deletion is quite different. The upstream region has ~28% GC, while the downstream region is ~45%. Do you think this difference in GC content is large enough to influence our crossing out events?

Thanks for any help again!