Hairpin template in a PCR

TR taskan4 at
Mon Oct 2 04:11:51 EST 2006

> Austin So wrote:
> the alternative is to use a primer with a very high Tm (65-70C), and use
> a 2-step protocol (combined annealing-extension and melt)

Sounds good. I could even use a 200-bp or so fragment, obtained by PCR
from the 500-bp fragment, as a primer, and the two-step PCR. Looks
that even the original 500-bp fragment could be used a primer, and
would be transformed into the hairpin in the PCR, then the Tm for the
closing of hairpin and for the binding of the primer would be the
same. However the other half of the hairpin will still be closer and
easier to find that the primer.
The problem with this approach is that I have in mind the application
of the same method to a heterogeneous population of DNA fragments, to
obtain an equivalent population of inverted repeats. Otherwise the
inverted repeat would be easier to generate by traditional cloning.
The heterogeneous population would have known ends:
I am limited to that ends for the designing of the primers, but it may
be possible to find a primer with Tm high enough.

> That is a good potential solution...and keep in mind that you need not
> limit yourself to phi29, but could just as easily use any polymerase
> with strong strand displacement activity, as long as you are able to use
> an appropriately designed primer in lieu of the random hexamers (the NEB
> catalogue has a nice table that you can look at).

I'll have a look, however I understand that Phi39 is the polymerase
with the highest known strand displacement activity and (i'm not
really sure about this one) the highest processivity. And these two
are important for this application.

> That being said, it is not clear that you will end up with a nice clean
> linear concatenate, since as the pol continues the hairpin will likely
> form immediately before being primed by any hexamers. In other words, I
> don't know if you will ever generate the linear molecule that you
> desire. I suppose if you used a specific primer with its complement, you
> could do it at an elevated temperature...

As I see the reaction, every new strand being synthesized by the pol
could eventually be displaced (slowly, step by step, starting from its
5' end) either by other molecule of the pol (or even by the same one
if there is only one working on the original single-stranded circle).
The strand being displaced should be more that 500-bp longer (in my
case, since I start with a 500-bp fragment) to be able to form a
hairpin, and before that happens the hexamers would be bound and Phi29
would already be working on the template, since the reaction is
carried out at constant low (I think 30ºC) temperature.
As you see I'm optimistic and we may give it try, as you said it
shouldn't cost too much time or money. The thing I´m worried most
about is the by-products that could arise in the amplification (I have
never done this before). I'm expecting kind of smear in the gel and
hopefully a stronger band of the right size.
Thank you again for all your help.

More information about the Methods mailing list