Gateway on BAC - Perform LR reaction on long constructs (Dec/07/2009 )

Hi all,

I'm encountering problems with Gateway reactions on long DNA constructs. I have:

Destination clone: a 30kb-long BAC, with an attR1-CmR-ccdB-attR2 cassette inserted.
Entry clone: derived from Invitrogen pENTR-1A vector (attL1-gene-of-interest-attL2). Regular size.

When I attempted to perform LR reaction to put the gene-of-interest into the BAC, I get VERY LOW transformation (electroporation) efficiency (but the colonies are positive). I think it's due to the problem of LR reaction. The reaction occurred but with a low efficiency.

Literatures I've ever read say that it's less efficiency working with long constructs, but don't give a solution to deal with it.

Are there any good ideas to optimize the reaction?

Thanks!

thomasz on Dec 7 2009, 12:53 PM said:

Either BP or LR reactions are not efficient at all with large constructs, I suspect the whole gateway clonase thing not to be efficient at all (just barely above ligase, which is weak), you might wanna include some sort of selection, like Kan resistance inside AttLx region, so it jumps into your BAC, though the reaction will get even more difficult if you add more bases.

Another extremely strong consideration is the efficiency of transformation on the receiving cells, how's it doing? 10^9, 10^10 should be ideal. Also, in your transformation efficiency test you should use something just as large (30kbp) to check. That's the most critical step I believe, the reaction might be going ok but the dna might not be going in.

Other tips are:

extra purify your entry and dest (recommended: any silica based dna purification scheme)
linearize you entry clone outside of the AttL region, even better if you could cut out the full attL region, elute and use that instead of the pENTRinsert
use sterile Tris-EDTA buffer to keep your purified DNA (or milliQ h2o, though it isn't as good as TE)
make a long LR reaction (16 hrs) at 20-25°C, by the book (4 uL LR clonase I to 20 uL reaction volume)
be as precise as possible in the entry-dest ratio (probably 1:1)


regarding the last consideration, I remember that int, (plus xis in the bp reaction) and ihf complexes are formed on top of dna with correct signal sequences, I suspect that if you add too much of attL entry vector, the enzyme will simply spend all it's time trying to cut out attL sequences, while the cutting and integration to AttR will be (let's say) less efficient. So it's a fragile multistep reaction, very forgiving for small inserts. But on long dna, you probably will get a lot of nonspecific binding of the proteins until it finds the right sequences, also there is the kinetics of dna bending and stuff like that which probably is affected by the dna size. They say it's like 20% recombination efficiency (or whatever), but we users know better.

Also to consider, though not in your case I presume, that too much dest vector might result in double transformants, in the sense that with an ultracompetent E. coli, you might transform your recombined dest vector AND a non recombined dest vector, the last one will kill the cell due to ccdB and you'll lose your clone. Or you might accidentaly transform the subproduct pENTR+Cm-cccdB. Either way your clone dies.


I'm doing a bp reaction (8.3 kbp PCR insert AttB and pdonr207 AttP), with a bunch of selections and it doesn't work at all. I believe it's the crappy cell transformation (tried almost everything, cell lines, rbcl, cacl2, 20°C, commercial prep, except electroporation). I've done dozens of pcr to pdonr to pdest reactions, very straightfwr work, got to protein expression in one month work, of up to 2 kbp insert. But the 8 kbp to 5 kbp has taken me more than 6 months (and more than 10 recombination attempts (4 transformations each) that's expensive stuff) with no results whatsoever.


Good luck, mate!

Rufflus

rufflus on Dec 23 2009, 10:57 AM said:

Oooops! I almost forget to check this post. My appology! Thanks so much for your looooong detailed reply!

My problem was somewhat solved later. The yield is about 1 colony per plate but that's enough for me.