How do I denature Herculase II enzyme? (Alternatives welcome) - (Nov/01/2011 )

So here's the issue. I'm developing a chemically modified aptamer, and the procedure requires a strand displacement step of a GC-rich hairpin loop. A few steps later, I need to do PCR. Unfortunately, the enzyme I'm using for strand displacement has been surviving and causing problems with primer dimers. I know it has to be the strand displacement enzyme because I've used hot start enzymes for PCR, and I've sequenced the artifact and confirmed it's a primer dimer. I've been using Herculase II (regular and hot start versions) for the strand displacement step. I'd previously used BST for strand displacement, but it wasn't a strong enough enzyme - I got replication of the loop ("laddering") in addition to the desired strand displacement. Even using manual hot start, there was some "laddering" observed by PAGE when I used BST. It was easy to get rid of BST, since between the strand displacement and PCR there was a selection step with boiling water elution. But Herculase can't be heat-killed like BST.

I'm currently looking into solutions, and a few ideas present themselves:

1) Denature Herculase. I've been looking for ways to do this, but so far my searches haven't turned up anything.
2) Perform a protein/DNA separation. But since I'm working with a small library that I'm doing selections on, I need to be really concerned with yield. Also, it's hybrid DNA: ds linked by the now displaced hairpin to highly modified ssDNA. Will normal separatory techniques work? I'm not sure. I can try though.
3) Switch to yet another enzyme. But what's a good one? I need an enzyme that excels at strand displacement of GC-rich regions and that can easily be heat-killed or otherwise rendered permanently inert. Is there a more appropriate enzyme I'm just not thinking of?

Thank you for your help.

How about a protease? Proteinase-K? You could inactivate it with PMSF prior to the PCR.

I'll try that. I also got advice from a post doc in my lab to try inhibiting it with BSA. He found it helped when an antibody was interfering with PCR. Mind you, an antibody wouldn't be degrading the enzyme like a proteinase would. I'm still puzzling out why it worked for him. But I'll try it.

Never worked with Herculase, but perhaps it works if you remove the Mg²⁺? I.e. adding sufficient EDTA.

hobglobin on Wed Nov 2 17:07:44 2011 said:

Interesting possibility. I'll definitely look into that. Question is, will Vent (exo-) also work without Mg²⁺? Then again, I could add a supply of Mg²⁺ for the hot start rather than enzyme. I'll need to switch to preparing my own buffer rather than using the commercial one, but it may be doable. But first thing's first: I'll need to figure out if Herculase II requires Mg²⁺.