PCR Mutagenesis

bhjelle at carina.unm.edu bhjelle at carina.unm.edu
Mon Nov 16 09:43:46 EST 1992


In article <Bxntyw.Fqu at usenet.ucs.indiana.edu> jgraham at bronze.ucs.indiana.edu (the End) writes:
>In <mxdq3j#@lynx.unm.edu> bhjelle at carina.unm.edu () writes:
>
>>I encountered an impressively high level of mutagenesis (14 
>>errors out of 411bp) in amplifying a segment of DNA from
>>a plasmid clone. All mutations involved T or A in some
>>manner. 11 were point mutations (transitions and
>>transversions), but 3 were 1 bp insertions or deletions.
>
>Very interesting and disturbing. We routinely generate templates for in vitro
>transcription via PCR using a protocol (Higuchi et al. NAR 16 (15) p.7351) 
>recomended by a well-known colleague.
>
>This procedure uses plasmid targets at 200-400 ng/ rxn. with dNTPs at 0.175 mM
>and a mere 16 cycles. The low cycle number is suggested in order to 
>reduce the chance of errors generated by Taq. I obtain about 1 to 5 ug of
>500 bp fragments from target plasmids around 3 Kb.
>
>In the NAR paper describing this technique, Krummel and associates report that
>they have not seen a single base change in preparing 10 different transcription
>templates in the 500 bp size range (p. 7359 line 14).
>
>Are you using a significantly lower dNTP concentration ?
>
I used 0.8mM of (each) dNTP, 4mM MgCl2 (these primers were optimized
at those concentrations) and 250 pmol of primer (each) in TV=125ul.
I got a high yield (5ug) of PCR product, but in back-calculating,
it seems that I would have gotten 250ug of product if all
dNTPs in solution were really "used up" in the reaction. So
I have a hard time believing that Taq ran out of dNTPs during
PCR. However, I did use 45 cycles for convenience.

At the suggestion of the P-E rep, I lowered the dNTPs to 200uM,
the Mg to 1.5mM, and did 10, 15, and 20 cycles to regenerate
PCR product. These parameters should all be lower to generate
less mutations. Also, I previously used 12.5U of AmpliTaq but
now lowered that to 2.5U. I used 1ug of plasmid template(!).

Hopefully I will be able to get back with results showing lower
levels of mutations.

Brian



More information about the Methods mailing list