need help for the transformation

MAEDA, Tatsuya, Ph.D. maeda at
Wed Aug 30 04:01:30 EST 2000

Here is my hypothesis.

Imagine what was happening in the last cycle of your (or anyone's) PCR.
Almost no replication, but just denaturing and re-annealing.  (You won't
get twice more PCR product by doing 30 cycles than 29 cycles.)  Therefore
almost any PCR product you got must be heteroduplex, each strand of double
helix having entirely different sequence in the middle but same sequence in
the ends.  By ligating them into the plasmid vector, then transform the
ligation product into the E.coli cell, each E.coli takes up a plasmid with
the middle heteroduplex region.  If you are lucky enough, E.coli repairs
the heteroduplex region of the transformed DNA into normal duplex before
the first replication separates it and fixes it into two plasmids with
different sequences.  Otherwise, you will end up with mixed plasmid
population from single colony.

The most serious part of the story is this odd phenomenon can occur not
only in the PCR reaction with random templates.  This was, and is, my
nightmare that I want to share it with members of the group.  

We know that PCR is an error-prone process and that mutation occurs at
rather high frequency.  Situation is much worse.  PCR products contain
mutations, yes, but they have mutations as mismatches.  Again, if you are
lucky, you have single plasmid by cloning the PCR product.  If not, your
single colony contains two plasmids.  Current sequencing technology doesn't
tell you whether the plasmid contaminated by, say, 5% of other plasmid with
one point mutation.  I have experienced this kind of problem several times.
 Now I always retransform the plasmid from the original transformant before
sequence confirmation.

I hope this comment might do something good to you all.

Tatsuya MAEDA, Ph.D.

Laboratory of Molecular Structure and Function,
Department of Molecular Biology,
Institute of Molecular and Cellular Biosciences,
The University of Tokyo

1-1-1 Yayoi,
Bunkyo-ku, Tokyo 113-0032

TEL: +81-3-5841-7820
FAX: +81-3-5841-7899
e-mail: maeda at


More information about the Methods mailing list