Hairpin template in a PCR

[TR]

Thank you to those who responded (see question and responses below).


I am going to try Betaine and the polymerases I can found in the labs
around, to see what happens, but I am quite worried after reading
Peter's response (not very optimistic, but quite well informed). I
jumped to PubMed to look for suppression PCR, and didn't like what I
found. Indeed it looks that my PCR is going to be a hard one. Well,
after the initial panic, I have been reading and came up with an
alternative solution. And again I would like to know your opinion
about this plan B:


A graphical version of the plan B is in
http://www.flypicture.com/bin/?id=qNj3lKrc

1.- Ligate the 500 bp-fragment to TWO different hairpin oligos, one
binding to one end and the other binding to the opposite end. This
would result in a single-stranded circle.

2.- Amplify this circle by Rolling Circle Amplification with Phi29 DNA
polymerase, so that it will transform into concatemers of
double-stranded DNA consisting of repetitions of the whole circle
generated in 1. In case you are interested, the original paper
describing the Rolling Circle Amplification is freely available in
http://www.genome.org/cgi/content/full/11/6/1095?ijkey=86fed9f82a6c8ddb8e90da5a9c4c7897798367ae

3.- Cutting with a RE in the sequence corresponding to one of the two
hairpin oligos would produce an inverted repeat of the 500-bp
fragment, with the other oligo in the middle.

Should I order the second hairpin oligo and the Phi29 polymerase right
away?, or am I missing something?

Well, I'll wait for your comments, and thank you again for the responses.

TR


Original question:



> > I am trying a strange PCR designed to transform a short DNA fragment

> > (500 bp) into an inverted repeat. Basically the idea is to: 1) ligate

> > the fragment to a hairpin oligo and 2) Set up a PCR with a single

> > oligo that binds to the opposite end. I have put and image to make

> > this clearer in http://www.flypicture.com/bin/?id=qNj0karb



Response 1:



> Won't work, can't work.  During the annealing step of the PCR cycle, all

> your single strands form up into neat little hairpins and don't let the

> primer bind, so the PCR fails.  This effect (suppression PCR) is regularly

> used by Clontech in their various selection/subtraction kits to *prevent*

> amplication of unwanted material.

>

> Why don't you just amplify your 500bp using an extended primer to

> introduce

> a restriction site at one end?  Cut it, ligate it to itself, and that'll

> give you the dimeric inverted-repeat product you're looking for.

>

> Peter



Response 2:



> I would be surprised to see that you get anything with Taq. When you

> think about it, a 50 mer duplex would melt at 80C or something like

> that, so you really need to find some additive that will split the

> duplex apart and allow a polymerase to plow through.

>

> You could also try 1M betaine, or even single strand binding protein.

> And add a whack load of primer. I think Invitrogen has some DNa binding

> proteins in their Accuprime Taq. You could also try Phusion or

> PfuUltraII to see if they will work since they have a built in DNA

> binding domain. Vent is apparently a strong strand displacer at higher

> temp and is less prone to slippage in the presence of secondary

> structure (I forget the reference).

>

> Cheers

>

> Austin