PCR:Sequence, or clone and sequence?

Charles Delwiche delwiche at sunflower.bio.indiana.edu
Mon Sep 21 14:40:05 EST 1992

In article <1992Sep21.101742.1700 at gnv.ifas.ufl.edu> afc at gnv.ifas.ufl.edu wrote:
>We have a question regarding experiment design.
>We are planning to PCR amplify and sequence the transcribed spacer between 
>the 18S and 28S rRNA genes from individual mosquitoes.  The purpose is to
>compare different populations, which may be subspecies or sibling species.
>We are considering two different ways of doing this:
>1. Sequence the PCR products directly.  (It is a little tricky getting 
>good sequence out of PCR products, but we have had success.)  This will
>give us a consensus of the sequence of the hundreds of individual genes
>from each mosquito.  Rare variants and PCR artifacts will not be seen.
>Common variants will give ambiguities in the sequence.
>2. Clone the PCR products and sequence several clones from each mosquito.
>This is obviously a lot more work, but it will pick up those rare variants.
>Unfortunately, it will also pick up PCR artifacts, and we will not be able
>to tell the difference.
[omitted text]
>Andrew Cockburn

Several people in our lab have been sequencing the internal
transcribed spacers ("ITS-1" and "ITS-2") for low level (i.e.
population and species level) *plant* systematics.  We sequence the
PCR product directly, as a double stranded product, and have gotten
excellent results with this approach.  We generally get clean
sequences if the PCR product is good.  In my opinion points one and
two above nicely summarize the advantages and disadvantages to each
technique, and indicate to me that you have a good handle on the
problem.  I'm with you -- sequence the PCR product and avoid the
cloning step.  This would *not* be my recommendation if you wanted to
study the spacer itself, for example if you wanted to study
intra-organismal variation in the spacer (which exists in multiple
copies), or in some related situations, but for what you are doing
direct sequencing ought to work beautifully.

I have another thought on the matter: if you clone and sequence you
will identify variants, but you will then have to determine what is
responsible for that variation: misincorporation of nucleotides durig
PCR, or real intraorganismal variation.  On the other hand, if you
sequence the double stranded product and find two or more superimposed
sequences, that would be very interesting, and would then call for
cloning from that PCR product and sequencing a number of clones from
it, but you would be doing that based on evidence for the presence of
a mixed population of PCR products.  If your mosquitoes turn out to be
similar to our plants, you won't find a lot of variation, and
sequencing the double stranded product should work well.  Sequencing
of double stranded PCR products can be a real pleasure, and can yield
sequence of comparable quality to single stranded DNA.

Good luck!

Charles F. Delwiche, Dept. of Biology        |"I see angels on ariels
Indiana University, Bloomington, IN 47405    |in leather and chrome..."
                                             |              James Adie

More information about the Mol-evol mailing list