Long PCR of genomic DNA

Bryan L. Ford fordb at bcc.orst.edu
Thu Apr 9 02:57:16 EST 1998

John Fellers wrote:
> I am trying to amplify up tp 10 kb of genomic DNA from tobacco.  Has
> anyone tried to amplify this length, and if so what were your
> conditions.  I am currently using a equal ratio of Taq and Pfu plus a
> modified 10X buffer.  Any help will be appreciated

If you wish to do this the "hard" way, that is by making up your own
long PCR enzyme mix, then please consult Wayne Barnes' original long PCR
article in the April 1994 PNAS first. Your apparent ratio of 1:1 is way
off. You need to have more like 60:1 or 200:1, that is you need only a
bit of 3' -> 5' exo activity to prevent stalling, but it is good to have
reasonable processivity, hence the desirability of taq. Also you must
find a buffer system that is consistent with both polymerases-- normal
Taq or Pfu buffers will not do this, but Stratagene can tell you what
ingredients will make a common buffer for these two. 

My first successful long PCRs were done shortly after Barnes' article
appeared and used Pfu exo minus with Pfu at about 100:1 in Pfu buffer at
one half the nominal 1x buffer concentration. Cycle parameters were
important to the functioning of these pioneering protocols, but I
believe the newer commercial protocols are somewhat more lenient is this
regard. But it any case you must clearly have long extension times, and
it certainly is helpful to subject the polymerases to the bare minimum
of total denaturation time (best accomplished in a Stratagene Robocycler
since you can set the denaturations for a very short time at 99
degrees-- so that the tubes are withdrawn just as they reach say 94
degrees). Another issue that Barnes discussed is the need for more
alkaline buffer systems to assure less depurination from the temperature
dependent acidic shift of pH. This last change may have percolated
through many commercial PCR buffer systems since Barnes' work.

There are doubtless by now several good long PCR systems available
commercially, I happened to try a couple early on and had exceptional
results with Boehringer's, for what it is worth, I now like to use
theirs for all PCR work regardless of the product length or fidelity
demands (I'm only a customer of BMB, no other relationship, and likewise
for Stratagene).

One last caution: you must have high quality template-- that is you
cannot expect to see long products from genomic DNA that has nothing but
short pieces of template. Column preps can help this problem a lot.
Related to this, I find that each and every freeze/thaw cycle will
reduce yield of >10 kb products substantially-- so if you must freeze
then it may be best to do it with LN2 "snap" freezing which allows a
glass transition rather than ordered (and apparently damaging)

Good Luck,

More information about the Methods mailing list