By far the most stable cloning vector I've ever tested is the linear "pJAZZ" plasmid (Lucigen Corp.). [Conflict of interest notice: I currently work at Lucigen and helped develop the vector.]

We've used the pJAZZ vector to clone di- and tri-nucleotide repeats, inverted or direct repeats of up to 40 kb, and highly AT-rich genomic DNA. It is very rare for cells to kick any parts of the insert out of the linear plasmid.

The multiple transformant colonies that you see likely began with just a single plasmid, but some population of the plasmid within the cell lost the insert. Subsequently, both forms of the plasmid are maintained in the cell. Eventually, you'll probably lose the larger plasmid from each transformant.

If you don't wish to use the linear vector, a BAC vector may be the next best alternative. I suggest using one that has an amplifiable copy number, to simiplify DNA preps.

Dear Brian,

pBeloBAC11 vector is suitable for cloning large fragments (up to 300Kb) and kept as single copy in E. coli. This vector is available from NEB for FREE.

http://www.neb.com/nebecomm/products/producte4154.asp

Here is the map of the vector

http://www.neb.com/nebecomm/tech_reference/restriction_enzymes/maps/pBeloBAC11_map.pdf

However, it it is likely that the fragment you have is expressing something that is toxic to the bacteria.

It is also likely that you have a unnoticed mistake of restriction enzyme choice. In another words, the map or the information you have on your fragment and vector may not be correct.

Best of luck!

I've never worked with such big plasmids. Yes, they probably are low copy. It would be great to find a way of including in your insert another selection marker, otherwise you'll always get back the empty vector. In one of my jobs I designed a way of joining together 3-4 overlapping sequences by gap repair in Saccharomyces cerevisiae. Then I made a plasmid prep that I electroporated into E.coli without problems. Selection both in S. cerevisiae (URA) and later in E. coli (Amp + NAT) ensured that I had the right constructs.

I used pPNT for my knockin constructs and the size is 15-17kb. it has a floxed pgkNeo. I cloned every piece into a separate plasmid and did three way ligation into pPNT at the end. I did found the plasmid was unstable when I tried to replace on piece - just my observation. So you may want to try to clone your DNA pieces in at once . I normally grow the bacteria in 1:1 LB/terrific broth - which helps for the yield.

Thank you all for your suggestions!

Ronald: I had briefly found the pJazz earlier in my searches. The main problem I have with using it is that it's linear, and I need my construct to be protected from T5 exo. I'll have to figure out if there is an alternate way, I was very intrigued by it.

Bekir: That...is amazing. I was trying to figure out how to obtain a BAC vector backbone from a library BAC that we have, but this will be much better. And...it's free? Awesome! With regards to the toxic properties, it seems that the cassette itself is not, but maybe when joined with the homology arms it is being expressed? It's a possibility that I was hoping the stbl or sure2 cells would solve. Also, the sequence is most definitely suspect, but it's rather difficult to sequence through the whole thing. If this takes much longer I may need to just primer walk the entire sequence :(

Francisco: I had actually just learned about ccdB today with my searching, perhaps I could clone something like that into the cassette first prior to this step. It's a good idea, thanks!

Ling: That is an interesting suggestion, from a brief search, it looks like terrific broth is used for high protein/plasmid expression growth. Do you know if this means higher per copy number or higher bacteria number? And any particular reason for a 1:1 mix of it?

Thanks again!

Hi Brian,

we are generating targeting vectors on a commercial basis. We regularly have targeting vectors that are in the range of 20 kb. In our hands that plasmid backbone isn't the problem, but the bacteria you are working with or the specific sequence that you are handling. For some reason it appears that bacteria don't like one or the other stretch of nucleotides and keep on cleaving it out. Try changing the bacteria you are working with to a rec minus strain and try cultivating the cells at lower temperatures (30°C). The vector backbone definitely isn't the problem!

Good luck

Stefan

Interesting, so you guys use high copy number backbone plasmids for targeting vectors? If I may inquire, what bacteria do you work with? I have tried stbl4s from Invitrogen, but have not yet tried the Sure line from stratagene.

We're using XL10Gold and if problems come up we go into B101.

These problems you're mentioning to appear, on a very low level so. By changing bacterial strains we so far always managed to circumvent or solve those problems.

The best circular plasmid is: Conditionally amplifiable BACs: switching from single-copy to high-copy vectors and genomic clones. Wild J, Hradecna Z, Szybalski W.

Genome Res. 2002 Sep;12(9):1434-44. You keep them at low-copy as a BAC and when you need a lot of DNA using relatively small cultures, you induce to high copy number. I have used it for plasmids sizing between 10 and 200 kb, in numerous publications. We even made an entire overlapping duplication kit in vivo in the fruit fly Drosophila using a library of 80 kb insert plasmids: A molecularly defined duplication set for the X chromosome of Drosophila melanogaster. Venken KJ, Popodi E, Holtzman SL, Schulze KL, Park S, Carlson JW, Hoskins RA, Bellen HJ, Kaufman TC.

Genetics. 2010 Dec;186(4):1111-25. doi: 10.1534/genetics.110.121285. Epub 2010 Sep 27. I think it is (one of) the best backbone(s) made ever so far.

Thanks for the references, I'll have to check them out. I was definitely noticing the extremely low yields I was getting from a normal BAC prep. I would just be worried that the last amplification would again destabilize my construct.

The best option is pBluescript with Takara long ligase overnight ligation. Use Invitrogen One Shot TOP10 electrocompetent cells. It have spent one year testing different products for cloning inserts up to 20 kb and believe me, this combination is the best! And here is my article about it: doi:10.1371/journal.pone.0047654

Either pCC1Fos or pBleoBAC11 can hold much mor then 20 kb, but much simpler would be to go for any vector with p15A ori. In my case the minimal backbone (just ori and resistance) could hold up to 25 kb.

Try first the promoterless backbone, to be sure that potential toxicity wont effect the subcloning. Due to your electroporation results it seems like you have a counterselection against your insert. Try SURE E.coli strain.

I'm also trying to clone a E.coli 17 kb plasmid DNA for sequencing. I have contaced differenr company but their price is too high more than 55k.CAN anybody help me and send me any vector that I can clone .