Long-amplicons and Taq

Cathal Garvey via methods%40net.bio.net (by cathalgarvey from gmail.com)
Wed Jul 6 07:26:55 EST 2011


(Crossposting from Google+)
I have a plasmid that's, say, 3.7kb long, and I need to amplify it with taq
polymerase (because that's what I have, and I'm too poor to afford better
right now). It's the length that's the main problem, although challenges
relating to the specifics of how to amplify it are also troubling me.

Specifically, I need to amplify a ~3.6kb plasmid *from another cloning
vector* as a linear molecule such that either amplicon end is complementary*
to the other, because this will enhance B.subtilis' ability to take up the
DNA and recombine it into a ring-shaped plasmid once more..or at least
enhance the plasmid's likelihood of copying itself successfully into a
ring-shaped plasmid by rolling-circle-replication before it is lost.

Taq seldom amplifies longer than 2kb, but there are voodoo ways to make it
do so. Among these voodoo methods are alternate buffers (apparently removing
potassium salt can improve matters?), different cycling conditions, and
different setup of the reaction.
I gather one of the big limiting factors is mismatches; taq extends
piecemeal, and if it detaches at a mismatch then there's no 3' end bound to
the template for the next enzyme to bind. So any method that enhances enzyme
fidelity should also enhance extension length.

Does anyone out there have tips on how to reliably achieve up to 4kb
amplification with bog-standard taq? I have a limited ability to make my own
buffers, and I can play with Magnesium and NTP concentrations easily enough.

*Anyone with information on how much complementarity is enough to induce
recombination in B.subtilis in this case? DNA is generally minced into
single-stranded DNA on uptake, and either end may suffer "nibbling" by
nucleases before circularising stably, so I'm imagining primers with a
10-15bp 5' "tag" complementary to the other end of the sequence.

-- 
letters.cunningprojects.com
twitter.com/onetruecathal
http://www.indiebiotech.com


More information about the Methods mailing list