Subcloning dirty PCR products

Zophonias O. Jonsson zjons at
Sat Sep 3 16:53:41 EST 1994

In article <34a0ilINN1k8f at>,
krasel at (Cornelius Krasel) wrote:

> Zophonias O. Jonsson (zjons at wondered why PCR products
> can be subcloned much better after proteinase K digestion (reference:
> Crowe et al, NAR 19(1), 184, 1991).
> The hypothesis offered by the authors is that the Taq polymerase sticks
> strongly to the DNA. Since agarose electrophoresis is not denaturing for
> proteins you would expect the polymerase to co-migrate with the DNA.
> Even denaturing of proteins with phenol seems to leave Taq polymerase
> on the DNA.
> --Cornelius.

Zophonias O. Jonsson (zjons at is not convinced !  :)

Taq pol is a polypeptide with calculated molecular weight of 93.91 KDa and
one would expect a protein of this size to slow down the DNA fragments
somewhat if it were bound to them.  The estimated net. charge of the
Taq-pol-polypeptide is -11 and that could compensate a bit for the size but
it still should slow down the DNA a lot, especially small fragments.  If
there were some variability in how many Taq pols each fragment would bind
one would expect a smear rather than a sharp band on the gel but PCR
products usually (altough not always) form nice bands on agarose gels.  One
could argue that the pol is probably bound to the ends of the fragments so
each one has exactly two of them.  That would be a nice theory wouldn't it?

Well... let's do a PCR.

We use 5u of Taq and amplify a 500bp stretch of DNA.  Our yield is 5ug
(I think this is realistic, even a bit conservative)

1 ug of 500bp DNA is approx. 3 pmol, so we have 15 pmol DNA 

Now the question is how much Taq do we have.  I couldn't find the specific
activity of Taq pol but if I remember correctly it is more processive than
Klenow so we can use the specific activity of Klenow as an estimate.  (The
danger here is that the unit definitions are not the same but they are
similar so the numbers shouldn't be that far appart)

Klenow specific activity: 20000 u/mg

		 5 u : 20000 u/mg = 250 ng

		250 ng : 93910 ng/nmol = 0.00266 nmol = 2.66 pmol

So we have less polymerase than PCR product unless the specific activity is
far off or there is an error in my calculations (I am like Taq pol rather
error prone when working under wrong conditions).  Thus, until someone can
find the flaw and show me that there is excess Taq polymerase in a normal
PCR reaction I am not convinced that this is the explanation.  
As I said in my last posting, the proteinase K trick has worked for me, so
I should just be happy and keep on digesting.  Long live molecular woodoo. 
But I am still childish enough to want to understand.....

Yours most sincerely


More information about the Methods mailing list