ligation/transformation: HELP ME PLEASE!!

saboteur at BORCIM.WUSTL.EDU saboteur at BORCIM.WUSTL.EDU
Wed Mar 1 14:01:06 EST 1995

>Would you believe, I am STILL having problems with the simplest cloning 
>exercise in the book?? I desperately need advice: in the past, I thought 
>my problem was trying to clone large inserts, but now that i've moved on 
>to a 1.1kb and a 2.7kb insert and my problems are still ever-present, I 
>am beginning to wonder... In the past, I have cloned into pBlue and bacteria 
>with great ease. 
>here is the story in brief:
>-- 2 HindIII fragments 1.1 and 2.7 --> genecleaned and ETOH precip. (others 
>in the lab have claimed problems with geneclean, either inhibitors 
>present for the ligation or transformation. I, personally, have never had 
>this problem, but the concensus is that it's the lot or it's old. I wanted 
>to be safe.)
>-- HindIII digested pBlue -->dephosphorylated-->genecleaned-->ETOH precip.
>-- ligated using brand-new BoeMannheim enzyme/buffer: 3:1 molar ratio 
>insert:vector, 10ul total (1ul ligase), 8 hours at 16 C.
>-- Ran 1 ul of ligation on agarose gel: different 
>conformations/smear/bands above vector size (I was told this is good: 
>it shows that the ligation was successful and circularized; thus, it runs 
>based on different coiled-ness); little/none discrete insert/vector bands 
>-- Transformed 3ul out of 10 (ligation rxn) into BOTH Sure2 cells and 
>XL1-Blue MRF': almost NO colonies (2-6) appear on the plate. PUC18 
>control looks great. 
>        -- I only mini-prepped the Sure 2 set (the XL1Blue will be done 
>tomorrow) and I obtained weird results: random, multiple bands. All 
>different for each clone. Perhaps I should hybridize to see if anything 
>is left from my original clone.
>Could it be a sequence problem?? After all, these are supposedly 
>5'/promoter clones that have an intron and exon present, based on initial 
>analysis. Could there be weird palindromes or repeats that the bacteria 
>doesn't like?? What part to you think is throwing me off?? Also, I just 
>did an experiment where I found that the gene is also slightly induced by 
>IFNB. I am suspecting a GAS element in the promoter, perhaps. It is a 
>far-fetched hunch, but could that cause a problem?? I am ignorant. I 
>don't know what to do now.
>Many thanks for any help you can give me.
>-- Heidi

Heidi, relax. You can work through this. First of all, you suspect that
there is an inhibitor of ligation and/or transformation in your geneclean.
Try: Geneclean an uncut vector and see if it transforms as well as 
before genecleaning it. Be cautious to use the same amount of vector in each
transformation. Next, try to see the effect of geneclean on recircularizing
of a vector cut with a single enzyme: Cut a plasmid to completion, heat kill
enzyme. geneclean half of it, and do a standard ligation on the other half.
transform. Based on your gel pattern following ligation, it sounds as if your
ligation might have worked. However, this is not a reliable way to determine
if your ligation worked as planned, only whether or not the ligase enzyme
worked. ie. this assay is only good for telling you that it definetely didn't
work (insert & vector bands still present). Also, it sounds as if you
set up a single ligase rxn for each subcloning. This assumes you know how
much DNA you started with, as well as your recovery efficiency. Why not set
up multiple ligations in parallel, varying the insert:vector ratio around
what you think they are. Finally, I always do my lig'ns in low melt (LMT),
which eliminates the need for genecleaning, and speeds up the process.
Run your gel LMT gel in TAE buffer (note, the TAE is important, as
borate will kill your ligation. Cut out your band with a fresh razor blade.
I usually try to minimize the time of UV exposure to only a couple of
seconds, and use longwave. After recovering the band in a 1.5ml tube, I
spin it to the bottom, add 50ul TE (we usually use 0.1mM EDTA, 1/10 of most
people) and 1ul 1M MgCl. Depending how thick your gel was and how much
extra stuff you cut out (avoid, if possible), you should have <100ul in
your tube. Heat it to 70C for a few minutes and mix by tapping, then put at
37C to set up your rxns. I usually set up a few lign's containing 0ul, 2ul,
5ul, 10ul of insert, and 2ul of vector (~20fmol of ends) in 20uls. The
equation for fmol of ends comes from Gibco, and goes something like this:

                           kb of DNA
ug DNA = (fmol of ends) * -----------

Add the ligase before the gel polymerizes. Depending on how important this
ligation is, I would suggest ligating at 14C o/n, although you could go for
a few hours at RT. Also, for blunt ends, double all [dna]. When you are
done ligating, melt rxn @70C, put at 37C, use 5ul to transform (quickly but
gently mix after adding to bugs, as gel will polymerize in cold).
Good luck.

Brett Lindenbach
Lucille P. Markey Student in Human Pathobiology
Program in Immunology
Washington University - St Louis
brett at

More information about the Methods mailing list