Cohesive end ligation of dsDNA to ssDNA

Emir Khatipov khatipovNO at
Thu Nov 29 21:29:02 EST 2001

I think I have a nontrivial question here for cloning gurus.
Would ligase (T4 or Ec, or RNA ligase) ligate the following molecules to
each other:
1) vector linearized with 2 different 6-cutter restrictases (4-base
2) single-strand (ss) oligo with ends complimentary to the sticky ends on
the linearized vector?
vectorxxx    <--tgca-random-oligo(50bp)-ttgg--> xxxvector
vectorxxxacgt                                                aaccxxxvector

I failed to find any information on whether T4 ligase strictly require 2 ds
DNA molecules with overhangs as substrates, although it is known to work on
blunt ds ends, albeit quite inefficiently. E.coli DNA ligase have been
reported to ligate ss oligos (as well as RNA ligase), but using these
enzymes is not the best solution, because they ligate oligos to each other
as well. Moreover, ss ligation is not really relevant to the problem,
because there are two 4-base sticky overhangs.

So the question is does the 3-strand ligation work, and if work, how

A standard approach is to use the following strategy, e.g.:

vectorxxx|acgt    acgtcgtaggccc-random-oligo(50bp)-aaagggccca
vectorxxx    tgca|tgcagcatccggg

 i.e. 3 oligos + linearized vector. Oligos are heated, annealed and
phosphorylated, and then vector and ligase are added. I cannot use this
strategy because 1) I need random sequence to start immediately after
ligation point, 2) there are no convenient restriction sites.

I am going to transform the resulting gapped molecule into E.coli, but this
is another story, and it seems that transformation should work.

I will greatly appreciate your feedback.


More information about the Methods mailing list