DNA polymerases for PCR-sequencing - (Aug/04/2009 )

Hi,

I'm trying to PCR amplify specific loci from human genomic DNA from a solid tumor for the purpose of mutation detection (by sequencing). I was wondering, can I use Taq for this purpose, or should I use a high-fidelity polymerase? If so, is there a particular one that you recommend?

Thanks!

Definitely use a high fidelity enzyme. My favourite (and imho the best) is Phusion from NEB.

I disagree. Taq is just fine, and probably easier to use. You don't care about the fidelity of the PCR reaction in this case -- single bases can be wrong on individual copies with zero effect on the sequencing results. You only care about the statistical average, which will be just fine. Taq is cheap, robust, and probably the ideal enzyme for this application.

I agree with phage434. Standard Taq only makes an error every 10,000 bp or so, so unless you are amplifying very large fragments you should have no problems using regular Taq.

What about if your error occurs very early on in the PCR? Wouldn't it then make up a big enough proportion of the amplicons to cause a problem?

If you are doing pcr on single molecules, that could be a concern. Almost any normal PCR starts with millions or billions of molecules of template.

That makes sense.
So for what applications do you really need high fidelity for then? Only those with a very few copies of template?

Any cloning reaction needs high fidelity. Sequencing samples the AVERAGE base composition at each position. Cloning selects a single molecule, and gives a sequence which is a duplicate of that single molecule. So, a population average can have the correct sequence, while any individual molecule is wrong in some position.

ok...but one cannot directly sequence the PCR reaction, right? I was told to PCR amplify, clone in e.coli (TA cloning) and then sequence the clone...so in such a scenario a hi-fi polymerase is advisable??

No, it is routine to sequence PCR reaction results directly, if one is interested in knowing the sequence of the original sample (as opposed to knowing that you have a correct clone of that sequence).