ligation problems

Thu Mar 16 12:12:44 EST 1995

In response to a question about blunt end ligation,
Alexander Kraev wrote:

>There is no real special technique, you just have to accept that it
is not going to be as efficient as any sticky end-ligation. All you need
for success is: phosphatased, gel purified vector; a lot of ligase ( 1-5
Weiss units in 10 ul reaction mixture) and an excess of insert. Typically,
20 ng of vector and 20-100 ng of an insert are ligated in 10 ul. With
average competent cells ( 10E7 colonies per ug ) this reaction should produce 
a total of 200 colonies, of which at least 30% are recombinant. These
figures in fact show that such a ligation is about 10E4 times less efficient
than an average sticky-end one. Hope this helps, 

There are some subtle issues concerning the vector:insert ratio that
I'd like to expand.  In the case cited, the vector and insert were
about the same size, so the molar ratios recommended above would be
1:1 to 1:5, which is about right.  It's the molar ratio that governs
the outcome of the experiment.  If you make the mistake of dumping
100ng of a small insert on 20 ng of vector, you may get into trouble
for the reasons that follow.  As you raise the molar ratio of
insert/vector, you create a higher proportion of multiple inserts. 
For a purified insert fragment, this will give some clones with tandem
inserts; but it will also cause loss of  cloning efficiency because
the molecules that ligate into an inverted repeat configuration will
choke the plasmid replication apparatus and fail to survive.  So as
you try to kinetically pump the reaction with higher insert
concentration, you'll reach a maximum and then start dropping because
of the latter effect.  If you're cloning out of a complex mixture,
you'll just create clones with a mosaic of fragments inserted  that
don't belong together.  This leads to all kinds of frustrations
downstream in your experiment; and when overlooked can lead to serious

If you want to pump ligation efficiency with higher DNA concentration,
I recommend that you raise both vector and insert and keep the ratio
at 1:1.  This can be accomplished artificially by adding a molecular
crowding agent like PEG.   Sometimes it helps to increase the insert
over the vector simply because people are notoriously optimistic about
estimating their insert concentration.  If you don't phosphatase the
vector (or if you phosphatase it unsuccessfully), then you'll have to
increase the insert:vector ratio starting at about 5:1 (molar) and go
up to compensate for the favored  kinetics of circularization over
intermolecular joining.  If you do a series of different insert:vector
ratios, then favor working up the colonies from the lowest
insert:vector ratio, because these are most likely to be single

Steve Hardies, Assoc. Prof. Biochem., Univ. of Texas HSC at San
Hardies at

More information about the Methods mailing list