Chromosomal recombination with BL21? - (Dec/25/2010 )

Hi there,

thanks for hosting me in this forum.

I'd like to ask if anybody has ever knocked in/out genes for any E. coli strain BL21(DE3), pKD46 mediated.
I have to insert in the chromsome a 6Kb regulon, but so far I have had no success using standard recombineering protocol.
I did not tested it in parallel with other strains (i.e. MG1655), but in K-12 and other B strains I had no problems obtaining what I was after quickly and efficiently.

I tried increased amount of linear DNA (up to ug), but no colonies.
Also, after recover, I subcultured my transformants in selective liquid medium (instead of solely plating), and the resulting population had severely impaired growth rate compared to the parental strain.

Additionally, I have the vogue idea that doing all my growing passages in chemically defined medium (CDM), somehow selects strongest cells for such a condition (since then I run bioreactor with CDM); this implies reduced transformation efficiency, but puts additional selective pressure towards my final condition.

Could anybody tell me if:
1) he/she has had any successful genes deletion/insertion in BL21(DE3) (or derivatives);
2) if there has had experience with other gene transfer solutions (i.e.: phage) and if it would be a better option; and
3) prove/disprove that growing bugs in CDM from single colony to large bioreactor leads to stronger phenotype for the condition of growth in minimal medium.

Many thanks for your support. Best wishes for the new year.

Miche

Hi miche,

concerning your questions:
1) Yes, i did successful recombineering in BL21(DE3), but very rarely with pKD46! I always experienced problems with that plasmid in combination with that strain. I now use the SIM-Plasmids from Donald L. Court, they work quite fine.
6 kb is a rather large fragment, you should use more DNA (i would suggest 500-600 ng). I also use LB-low salt for the preperation of the cells and just use 1mm gap cuvettes for electroporation. What loci you are targeting? Are you sure that these genes are not essential under your conditions?

2) Yes, i have. We have been successful in doing integration in MG1655 and transfere the created traits by P1 transduction in BL21(DE3). This worked good for us as well. But you must bear in mind that beside your trait chromosomal DNA from the K-strain is transfered in the B-strain. This can modify the genotype significantly.

3) I do not really get your assumption for this point? Do you mean you are going to grow your cells for your ions engineering in bioreactors in minimal medium? ...or what do you mean? Growing cells in minimal medium will need more genes in a functional state that in e.g. LB medium. Therefore minimal medium is a real burden for the cells when they have been stressed before. Maybe you can explain you approach in more detail?

Good luck!
Regards,
p

miche on Sun Dec 26 03:28:56 2010 said:

Hello pDNA and everyone,

thanks for your reply, you gave me very useful hints.
I was not aware of the existence of pSIMs plasmids, but they seem to overcome my limitations. Here’s explained my poor efficiency with simple pKD46 transformation into BL21 (at least one order of magnitude less than with other ECs).

1) My KI is directed towards the Lac operon (lacZ promoter and bits of lacI-Z genes). It was easy enough to find homology regions from commercial plasmids (i.e.: pCR2.1); anyhow similar constructs in the other ECs worked fine. That’s why I was puzzled about BL. I have other vectors targeting other non-essential chromosomal genes, but it is irrelevant at this stage.

2) I agree concerns about DNA being transferred from one strain to another. However, the fragment I want to recombine comes from another strain than K12 or B and simply seems to work fine.
I never quite understood phage transduction (no practical experience), but to me looks like general pieces of chromosomal DNA from the donor, site-specific recombine into the recipient. As long as it is a site-specific insertion (i.e.: I target where the fragment of interest goes) I have no objections. Then the fact of going into another building to do transfection is sufficient reason to let me pursue the option pSIMs.

3) Sorry to have been cryptic.
My end point is bioreactor in CDM (chemically defined medium). I vaguely believe that growing cells from scratch (single cells, preparation of recombinant cells with molecular biology) in defined conditions, somehow selects (or adapt, as you like) the most amenable cells for the given condition. On the other side, I have the concern that preparing my clones in generic medium, could drive the required insertion into a cell that for some reasons (luck?) is weaker than the other, i.e.: unfit to grow on CDM, which would void any improvement obtained with the insertion of foreign genes. Therefore, I routinely grew from glycerol stock and prepared my electrocompetent cells into CDM. The price to pay is a drop in efficiency, that I would normally accept in other ECs given that enough positive recombinants were obtained (5ish per transformation). Unfortunately, this becomes unacceptable with BL21.

Playing around a little bit more with BL21 with pKD46 (sometimes in silly ways that I won’t bother reporting), I have obtained BL21 growing in CDM under my conditions, but with growth rate seriously impaired.
I found in the literature that some recombinants of BL21 have auxtrophies for some AAs and I wonder this is the case, since passaging the mutated colony into LB restores the expected growth rate.

The pivotal question I’m trying to address at the moment is whether BL21 is really a superior strain for protein production or it has just ‘been lucky’ to have T7pol and few useful deletions (lon, ompT, etc). Therefore, I’m attempting to transform several different plasmids into few EC strains working under the conditions I had set (this latter comes from the genome). As you can see I still have not obtained satisfactory BL21.
Have you ever tried another promoter than T7 to express anything in BL? Are you aware of anyone who did it?

I apologize for the terribly long post.
I wish everybody all the best for the new year.

Cheers,

Miche