TA-cloning heteroduplexes?

marc crepeau mwcrepeau at hawaii.rr.com
Wed Jan 25 17:45:20 EST 2006

I'm trying to gather sequence of a nuclear intron in a large population
of a diploid marine organism.  Because there are frequent indels I
can't direct-sequence all individuals.  For individuals with
heterozygous indels my strategy is to TA-clone the amplified intron.  I
then pick colonies, suspend them in a little water, and use that as
template for PCR with M13 forward and reverse primers.  Then I clean up
the PCRs and sequence them.  I screen colonies in this manner until
I've recovered both alleles.

The problem is I'm finding a high frequency of colonies that give mixed
sequence downstream of the indels, same as I would get from
direct-sequencing the amplified fragment.  I would expect to get some
small frequency of clones which experienced two independent
transformation events and therefor contain two populations of plasmids,
one with each allele.  But I would think that would be a fairly rare
event, and I'm seeing a lot of mixed sequences.

This makes me wonder if my PCR might be generating heteroduplexes in
the final melt/anneal cycle?  If such a heteroduplex was ligated and
transformed would the bacterial DNA-replication machinery resolve the
mismatches by creating one sister plasmid with one allele and one
sister plasmid with the other?  Or would only one allele be preserved?

If it's possible that I'm cloning heteroplexes, are there certain PCR
conditions that favor or disfavor heteroduplex development?  Naively I
would assume that continuing the PCR until the dNTPs and primers become
limiting would favor heteroduplex formation.  So maybe reducing cycle
numbers is advisable?

All comments welcome!

More information about the Methods mailing list