Pfu polymerase in PCR

Colin Dolphin c.t.dolphin at
Wed Feb 12 14:06:03 EST 1997

In article <A2E$4FAI4DAzEwHz at> Dr. Duncan Clark,
duncan at writes:
>Subject: Re: Pfu polymerase in PCR
>From: Dr. Duncan Clark, duncan at
>Date: Tue, 11 Feb 1997 09:38:16 +0000
>>In article <fof1-1002971723410001 at>,
>"Frank O. Fackelmayer" <fof1 at> writes
>>Pfu is a proofreading POLYMERASE. It will NOT degrade your DNA in presence
>>of nucleotide triphosphates (this is why you should add the polymerase
>>last). Only after misincorporation of a nucleotide, Pfu will remove this
>>nucleotide and then continue polymerizing. 
>Not quite true. Unannealed primers are single stranded and even in the
>presence of dNTP's the 3-5 exonuclease will attack them. When they are
>annealed as a duplex in the prsence of dNTPs polymerisation occurs and
>the exonuclease is inhibited from action but only on the duplex.

I¹ve had lots of problems with Pfu - my feeling is that, as Duncan says,
although the polymerase activity is preferred in the presence of dNTPs
surely when amplifying from a rare target (genomic DNA say) then for the
first few rounds, or perhaps more, the primers are going to be in excess
and as there¹s not going to be a lot of primer annealing going on then
there¹s not going to be much polymerising either. Therefore, for these
initial rounds there might be a significant proportion of Pfu hanging
around not doing any polymerising but quite happy to use it¹s exo¹
activity to chew away at your primers resulting in little or no product.
(This does happen as PFU is very good at correcting your
mutation-containing primer if your trying to use it to do PCR-mutagenesis
resulting in all wild-type products.) 

Although I¹ve given up with neat Pfu I do add it to regular Taq which
invariably produces PCRs with greater yield and allows me to amplify
longer sequences than with Taq alone. I wouldn¹t be surprised if the
sequence fidelity is also on a par with PFU alone (althugh I¹ve not
tested this as I can¹t amplfify with Pfu alone !)


More information about the Methods mailing list