Taq vs Phusion polymerases - specificity - (Nov/26/2012 )

Hello all,

I have been struggling in the last couple of weeks to amplify 2 sequences from cell isolated gDNA (both sequences are between 75-80% GC rich; 700-900bp long). I had lots of problems with unspecificity with both Taq polymerase and Phusion polymerases. I have tried 3-5% DMSO, GC buffer, temperature gradient for the annealing temperature of the primers, two sets of primers...non of these worked and I got lots of unspecific bands, but none of the expected size of my amplicons. Then, last week i tried Taq+2M Betain and it worked like magic! I was really impressed by the difference in specificity between Taq alone and Taq+Betaine and both sets of primers work fine with both sequences. So far so good, but I tried 2M betaine with Phusion polymerase and it was the same mess as without betaine - lots of unspecific bands. I need the amplicons for sequencing, so i would prefer to use a high fidelity DNA polymerase. Has anyone experienced a similar problem? Do you have any suggestions how to improve the specificity with Phusion? Is unspecificity a common problem with high fidelity polymerases in general, or should i try another polymerase, such as Pfu? Alternatively, i guess i could just take several replicates and do the sequencing with Taq - it would be highly unlikely that Taq would incorporate the wrong nucleotide in exactly the same location between 4-5 replicates. But I am also curious to understand why would high fidelity polymerases be more unspecific than TAQ?

I'd try to redesign the primers with as high annealing temperature as possible (or trying out the existing with higher temperatures). Also decreasing the Mg²⁺ concentration improves specificity.
Anyway with such a relatively short sequence I'd not care that much about the error rate of the polymerase, as it is anyway very low and because you can or even have to repeat it (e.g. sequencing from both directions to get the whole sequence). By comparing sequences you can easily find the possible errors, if there are any.

I think you have discovered that, through lots of work, it is generally possible to make bad primers work. Life is too short for this, however. In general, primer redesign is faster and more rewarding than spending a lot of time optimizing the PCR reaction with bad primers. On the other hand, betaine is definitely my go-to reagent in making poor PCR reactions work. As you point out, it often works like magic.

The problem is that both sequences (as i want to keep them below 1kb) are really hard to design primers for, as they are extremely GC rich in most areas, so I am pretty sure the primers I have designed are just the best possible for these specific regions (they do not form any dimers, loops, etc.). I might try increasing the annealing temperature even more and playing dropping the Mg concentration, just to see what happens. Thank you both!

Are you using the high GC buffer with Phusion?

Yes Veteran, it doesn't work, i have tried supplementing it with DMSO up to 5% and with Betaine, but it doesn't work. I am planning to try with ethylene glycol and 1,2-propanediol.

TMAC (Tetramethyl ammonium chloride) is also worth a try if you want to work with such specificity enhancing additives.

I work with Phusion, but I have to tell you it is not my favorite polymerase. We had difficulty amplifying our products for a while, but recently manged to optimize our reactions. We don't use GC buffer, because even for GC rich areas HF buffer worked better. We also use 2.5 ul of DMSO in 50 ul reaction, use up to 15 ul of cDNA (more than recommended), use 2.5 ul of 10 uM for each primer, use 1.5 ul dNTP 10mM, and 1 ul stock MgCl2 from the box, and 1-2 ul polymerase. Also we add water before adding DMSO. We like it this way, though it might not have any effect at all.

but one of the most important things was the annealing temperature. According to the manual you must keep your annealing temp at the Tm of your primer, or if the primers are longer than 20 mer you must increase the temp by 3C...but we found 3-4C lower annealing temp works even better than recommended. Also, for longer PCR products we keep our time at 40s/kb. We have got our PCR to work with 15s/kb too, but we prefer 40s/kb, although your products are much shorter than ours.

It is certainly true that Taq + Pfu mixtures are much more forgiving than Phusion. I have yet to do a systematic comparison, but have had consistent good luck with the new NEB Q5 polymerase, which in (uncontrolled) tests seems to give results more easily than Phusion. I use their 2x master mix (not the high GC version).