Recombination due to overselective plate? - (Jan/19/2012 )
Hi,
I just transformed a commercial neo/kan vector containing my gene's ORF and 8 out of 10 colonies I picked recombined the plasmid at the (presumably, I sequenced only 4 of them) same spot, cutting the gene whole out. The other two I'm not sure of yet, but they show normal restriction pattern.
The plasmid is 8.3kb (classic, not viral or anything) and contains full-length human kinase OR. There were some rumours that bacteria cut the gene out, from other labs, but I don't completely trust that for some reasons I'm not going to disclose.
Now the plasmid manual states it should be selected by 25ug/ml kanamycin, but I had plates with 50ug/ml (the liquid media for minipreps was 25ug/ml), they were bit older so I thought that wouldn't matter that much. The transformation efficiency was notably low.
Could the double-dose of Kan in plates caused the cells to recombine more then they would normally, because of the plasmid size and selection pressure?
(Given those two clones are fine I need to do a mutagenesis, so I need to troubleshoot and in case the recombination was caused by the inherent features of the gene in question I would probably try rec- strains, but I don't know how compatible these are with the mutagenesis kit.)
Thanks.
No, it should have little effect. It sounds as if your bacteria are actively selecting against the presence of this gene, both from the low transformation efficiency and from the sequenicng of mutants. You could try putting it behind an inducible promoter such as the T7 promoter and only express it if you need protein. You may need to find a plasmid with low or non-existent transcriptional read through. Orienting the gene in the opposite direction sometimes helps.
Thanks for the reply.
It's a mammalian vector, with CMV promoter. I need it to express the final mutants in cell culture, it can't be opposite direction. Also it's difficult to play with the insert as the vector is from a commercial system and I would like to transfer the gene to the same type vector with different mammalian selection marker, also from the company. It's no fancy cloning, just mutagenesis and experiments on stable transfectants.
If there is 20% that's OK (we'll see) that's still doable, for re-transformation. But I can't predict how this efficiency would drop after the mutagenesis, never done that.
So you think rec- (I have here Stbl2) won't help? There could be selection against my gene, but there has been like dozens papers published with this (or mutant) gene in cell culture expriments, and they had to clone it somehow, so either the selection is not that severe or there is a different player that can be put out of the game.
Trof on Thu Jan 19 14:43:44 2012 said:
Thanks for the reply.
It's a mammalian vector, with CMV promoter. I need it to express the final mutants in cell culture, it can't be opposite direction. Also it's difficult to play with the insert as the vector is from a commercial system and I would like to transfer the gene to the same type vector with different mammalian selection marker, also from the company. It's no fancy cloning, just mutagenesis and experiments on stable transfectants.
If there is 20% that's OK (we'll see) that's still doable, for re-transformation. But I can't predict how this efficiency would drop after the mutagenesis, never done that.
So you think rec- (I have here Stbl2) won't help? There could be selection against my gene, but there has been like dozens papers published with this (or mutant) gene in cell culture expriments, and they had to clone it somehow, so either the selection is not that severe or there is a different player that can be put out of the game.
You should try stbl2 or stbl3 and incubate your plate at 30 deg C instead of 37 deg C. Also shake your miniprep or maxiprep cultures at 30 deg C. You will see that it helps a lot! I work with viral vectors and often get recombinations due to the LTRs. Some constructs are just impossible to get.
The remaining two clones had both small deletion just before first codon.
As far as I was able to identify the sequence of CMV promoter (I blasted it with CMV), it is further upstream. But seems logical that this affected probably the transcription start or something very crucial.
Now I have no experience in vector engeneering and I have little time.
Is there some relatively simple change I can do to make me a mammalian expression vector, that would not express in bacteria?
Small change in sequence that would keep it expressed in eukaryotes but not E.coli?
Competent cells that wouldn't express from CMV promoter?
Different vector (which one)?
Or more basic is there a tool for analysing the vector sequence and identifying important elements? Or something like "express your protein from the scratch for dummies" 
I was thinking all kinds of variants keeping a kind of cripled vector for the manipulations and make it expressible only as a last step, but that last step is still a problem.
Still, the Origene was able to produce the vector, they have to do it somehow. But I have a feeling they won't tell.
I'd like to hear your opinion, or any comment, because I have to solve this quick or reconsider other options.
In the meantime will try new plates just in case and rec- cells, but since the problem is AFAIK not primary in recombination I don't expect much from that.
concentration of antibiotics can have an impact ...for details take a look at this paper.
what commercial vector you are using? ...and can you give the detailed description of the strain you use for cloning?
you have to rule out that there is no bacterial promoter that allows expression of your gene of interest (e.g. lac, T7, ...).
Using a lower temperature for the outgrowth after electroporation is for sure a good idea (30°C or even lower) ...if Stbl cells won't work try using SURE cells ...maybe your problems are due to Z-DNA or inverted repeats.
Good luck!
Regards,
p
This is my vector.
I'm using normal TOP10 from Invitrogen, but I used the last batch, so I have to look what we have in the freezer now, XL-1 Blue I think.
CMV promoter used unfortunately does express in bacteria a bit.
I looked closer into the vector sequence.
It seems that there is CMV promoter, after that a T7 promoter (TAATACGACTCACTATAGGG), and then RBS (GAGGAGA) and Kozac (GCCGCCGCGATCGC) sequence (doesn't fit the concensus, I don't know why). And from the cloning scheme it shows there are three aminoacids Ala-Ile-Ala added before my gene, I heard that's done sometimes in expression vectors (why?).
But.. the translation doesn't start at ATG?
The whole start of the sequence looks like this (red - end of CMV, green - T7, orange - RBS, violet - Kozac, underlined - translation, bold - my gene's first codon):
C GCG ATC GCC ATG
The two clones that don't delete the whole gene, delete either part of Kozac (9bp) or the three added aminoacids (9 bp). I really don't know if tha later is important for mammalian expresion.
to me your gene seems to be toxic at high copy numbers ...you can try using an strain like this ...that lowers the copy number or try growing your strain at very low temperature.
Regards,
p
I may have some hope 
Those 9 bp misteriously missing from my clones (those that didn't recombine) may be only error in sequence and scheme they published.
In their sequence there seem to be doubled the cloning site, which is not very possible if it was used for cloning and shows a translation start with alanine, which I really doubted from the first time I saw it. I wrote them to confirm, but maybe my clones are actually OK 
XL-1 blue cells on 25ug/ml kanamycin grew like hell and made huge colonies, roughly half of it was recombined too, but I could select the right ones by cracking.