How to remove the background and unspecific bands - (Aug/03/2011 )
Adrian K on Thu Aug 4 19:02:27 2011 said:
I had just missed out that you are actually using verbatim polymerase, which is a very high fidelity and fast polymerase with proof reading ability. If I guess correctly, you are using a thermalcycler with low ramp rate (~2C per second heating, ~1C per second for cooling). This is because your ~2kb band appears on the higher temperature rather than the lower one in your first gel, and since you using a low ramp rate machine, the annealing time and the extension time for higher temperature will be longer, due to the ramping which gives ~30 seconds longer for higher temperature gradient to maintain in place (i can elaborate further if my explanation sounds greek to you due to my bad expression).
Based on my experience using KOD polymerase (similar polymerase like yours), Such fast and high fidelity proof reading polymerase always gives me a hard time in optimization, really. This is because it doesn't really behave like normal Taq polymerase. Even under same annealing condition, Taq polymerase gives me my desired band but not in such high fidelity enzymes, and I use such enzymes only when I try to clone a large fragment (more than 2kb) and I don't want to wait for long time, else I will just stick to normal Taq. I have to reduce my extension temperature to 68C for optimal proof reading, and if I'm using a slow ramping rate machine, I need to reduce my extension time up to 5-10 seconds, to reduce the larger artifacts, and rising the annealing temperature up to 63C. Since your fragment is just 700bp, use a normal taq polymerase with Pfu blend will do.
Also, based on your label in your last gel, you are comparing the "GC-buffer" and "non-GC buffer" result, instead of "with/without DMSO", correct me if I'm wrong. And, according to the product insert http://www.sorvall.com/eThermo/CMA/PDFs/Product/productPDF_52581.pdf
the initial denaturation is 95C and denaturation is 98C. Is there any reason for you using 94C?
Thank you Adrian for explanation.
Recently our lab change to use Verbatim for HF PCR because of financial reason. You are right this enzyme probably is too fast to control PCR reaction, and resulting in unspecific bands. Because I will use a mutagenic primer afterwards, I wonder that Taq enzyme will influence the high fidelity or not.
As far as the annealing temp I used, this PCR programme works well with my collegues.
Biogareth,
How many cycles you used to run??
Try to make it less cycle....if u doing 30cycles, try make it 25...it can reduce noise 
or try to extend the denaturation time (maybe it helps)
I dont like to suggest this, but maybe you also need to redesign primer to get more specific one
Regarding to pcr your purified gene, did u check the band (after you purifying your gel) before run pcr? i am afraid, some of 500bp band was cut and and got purified too...
I would gel extract both bands TA-clone them, sequence and then decide what is going on. May be everything is fine with your PCR but you are dealing with a heterozygous deletion... Or if the second band is completely off target you should just redesign one of the primers... Sequencing will make it all clear!