Will ligase buffer affect polymerase fidelity? - (Jun/22/2009 )

I'm trying to construct a deletion in a bacterial gene. To do so, I'm amplifying a region of the gene upstream of the region I want to delete, and a region down stream. I've added EcoRI sites to the ends of the PCR products (3' end of upper piece and 5' end of lower piece). I then digest the products with EcoRI and ligate the pieces together. Following ligation, I use the ligation product (heat inactivated) as template for PCR using the outer primers from the original 2 reactions to amplify the deleted gene product. The whole process works pretty well, but after cloning the final PCR product and sequencing it, I've got 4 C>A transversions in the sequence. The size of the piece is a little less than 2 kb (about 800 bp for the part upstream of the deletion and 1.1 kb downstream) and I'm using Phusion polymerase from NEB.

My PCR conditions are:

50 ul reaction
0.5 uM of each primer
1.5 ul DMSO
1x HF buffer
1 ul of the ligase reaction as template
don't remember concentration of dNTPs, but pretty standard concentration
water to 50 ul


The point mutations really didn't affect my product, 3 were in non-coding regions of the sequence, and the 4th was in the wobble position, and didn't change the amino acid, but it's bothersome that there are that many mistakes in such a short sequence. The only thing I can think of is that the ligase (T4 from NEB) buffer somehow is affecting the fidelity of the polymerase, even though it's essentially a 1/50x concentration of the buffer in the PCR tube.

Any thoughts?

Thanks.

Why are u adding DMSO to the PCR?- that is ur culprit. Phusion HF works pretty well but the kind of error rate u are seeing would be unbelievable even with Taq for a 2kb amplification. get rid of the DMSO, it lowers ur annealing temperature and increases mismatches and should not be used but for really hard to do PCRs. If that is the case with your PCR, even GC buffer (other buffer sold with phusion poly.) would work better than DMSO- bear in mind that GC also increases error rate.

I've used DMSO in almost every reaction using Phusion and haven't had a problem with that kind of error rate. I don't remember why I started using it, but I think it's suggested in the book that comes with Phusion. However, I sequenced the PCR products I suspected to have the errors, but the PCR products were perfect, meaning Phusion didn't introduce any errors to the sequence. The errors are either something in the sequencing or something introduced by the E. coli strain used, which to me is a stretch. So I'll be resequencing the plasmid insert to see if the sequencing was the problem.

Anyway, the answer would be that the ligase buffer doesn't affect the polymerase.

I will keep the DMSO Phusion in mind for any future problems, though.