Wayne M. Barnes
Department of Biochemistry and Molecular Biophysics, Washington University School of
Medicine, St. Louis, Missouri USA and DNA Polymerase Technology, Inc., St. Louis, Missouri
wayne@barnes1.wustl.edu and barnes@klentaq.com
In 1992 I, and I assume others, were working on mixing and fusing various domains of different
DNA polymerases, in order to have the advantage of long reads and/or high fidelity, combined with the
robust reliability of Taq DNA polymerase. It was assumed that processivity (long "hang time", staying on
the DNA for long periods and long distances) was key to long PCR. That may yet come true some day, but
so far, the opposite is true. The Klentaq1, an N-terminal deletion of Taq DNA polymerase, is even less
processive than the full-length Taq, yet it was used to set the record of 35 kb in 1993. Ironically, less hang
time was found to allow other components of an enzyme mixture to access the growing chain to repair or
edit problem DNA molecules that were limiting extension -- those with mismatched bases at the 3'-end.
In this chapter I will use the acronym LA to mean Long and Accurate. Long being more than the
ca. 3 kb limit of efficient PCR without a mixture of enzymes, and accurate to mean high fidelity, since the
long PCR mixtures were found to simultaneously provide some 10-fold fewer mutations in the PCR
product (Barnes, 1994). I will also use LA as a suffix to mean a mixture of DNA polymerases, the major
one usually being Taq or Klentaq1 (which have no 3'-exonuclease proofreading activity) and the minor one
is an archaebacterial DNA polymerase such as Deep Vent, Vent, or Pfu (Barnes, 1994). TaqLA means Taq
DNA polymerase mixed with a small amount of Deep Vent or Pfu. TaqLA is sold under various names,
such as Expand, Elongase, ExTaq, TaqPlus, Accutaq, etc. KlentaqLA means Klentaq1 with a similar low-
level partner of proofreading DNA polymerase.
Other components are being introduced to improve LA PCR. Among these I mention one enzyme
component and one chemical component.
(1) dUTPase. Stratagene (Hogrefe et al., 2002) has discovered that a buildup of deaminated dCTP ,
namely dUTP was causing the incorporation of uracil into DNA. Uracil in DNA is a disaster for the
archaebacterial component of LA PCR mixtures of DNA polymerase, since the archeabacterial DNA
polymerases (such as Pfu) bind almost irreversiby to dU-containing-DNA under in vitro conditions (Lasken
et al, 1996). A cure introduced by Stratagene is thermostable dUTPase . This enzyme is now included in
some of their PCR enzyme mixtures, in addition to the DNA polymerase components with low and high 3'-
exonuclease.
(2) Betaine. Although introduced for high GC targets, I find that betaine never hurts, and usually helps,
even for long PCR up to at least 20 kb. It is included at surprisingly high levels, such as 2 molar final
(Baskaran et al., 1996; see tip #4 below).
Since the introduction of mixtures of DNA polymerases (Barnes, 1994), most PCR reactions, of
any length, have improved in reliability and in yield of product. Nevertheless, some amplifications do fail.
When I ask to see the detailed conditions for a failing PCR, I usually find that the experimenter has broken
about half of the following "rules" that I have learned the hard way. If you find yourself in column B on
one or more points, consider trying column A.
Warning: Your reactions may improve so much that they will over-cycle, causing the appearance
of a very fat product band, and visible side products that only become visible when a (finished) reaction
continues to cycle after the main product band is fully formed. For a maximally clean signal, you may have
to add less template and/or use 5 fewer cycles than you used to use with less efficient amplification.
DO DON'T
LA-PCR Tip #1 Use too much DNA template.
Use a maximum of about 1 ng DNA 1 ug is way too much template per ul of PCR reaction.
More is not always better. Something else must be wrong if you are tempted to use more than 1
ng template per ul of reaction. For plasmid as template, much less is best: Never use more than 0.1 ng per
ul, and usually use only 1 ng in the whole reaction.
DO DON'T
LA-PCR Tip #2 Use plain Taq.
Per 100 ul of PCR reaction volume, Instead, use a mixture
use about 0.1 ul of KlentaqLA per kb of enzymes licensed under
of target size. Above 10 kb target US Patent 5,436,149.
size, use a range of 1 to 1.3 ul enzyme. These catalyze better PCR
even for short targets.
I am biased in favor of my original mixture used to discover long and accurate PCR (Barnes,
1994), namely KlentaqLA, which is a mixture of Klentaq1 and Deep Vent. I find that this mixture gives
consistently more product than mixtures with full-length Taq as the major component. Klentaq1 is an N-
terminal deletion of Taq DNA polymerase, and it has slightly higher thermal stability (Barnes, 1995). I
believe the increased yield with Klentaq1 is because the 5'-exonuclease (deleted in Klentaq1) can
negatively impact the product at high concentrations. I say "at high concentrations" because the 5'-
exonuclease actually has poor activity on the end of double-stranded DNA; it is optimized instead for
forked structures, such as the structure encountered during nick-translation.
Even I find that, for some cases such as long RT-PCR, a TaqLA (mixture) product such as Expand
sometimes works better.
DO DON'T
LA-PCR Tip #3 try less Mg than dNTPs.
Consider Mg++ concentration as the Don't add dNTPs without their
excess over the level of dNTPs. After "own" Mg in their stock.
上一篇：Standard LA PCR Protocol 下一篇：LA PCR Protocol (Revised from Takara Shuzo Co., Ltd.)
共6页: 上一页 1 [2] [3] [4] [5] [6] 下一页
