optimizing overlap PCR, ligation-free cloning

Maximilian Haeussler via methods%40net.bio.net (by maximilianh from gmail.com)
Tue May 29 08:56:12 EST 2007

Hi Neeru,

no I didn't get any answers to my questions about the maximum yield of
a PCR and how to achieve it. Having asked around in the lab, I think
now that I won't get 5 ug anyways from a PCR tube of 100-150 ul.

So I've come to accept maxipreps instead and started playing around
with PCR-induced cloning where the product of the fusion reaction is
giving directly a plasmid that can be transformed. Remind me in a
couple of weeks and I can tell you more. (See also [5] and [6], if
anyone has any experience with "ligation-free" techniques , I
appreciate any criticism or weak points of these strategies)

I try to give more details here about the Fusion PCR, though it is
simpler than it sounds here:

The idea to join fragments A and B is to design primers at the end of
A that overlap the beginning of B.
So you're doing three PCRs: I (with special reverse primer) to amplify
A and II (normal primers) to amplify B. PCR III will then join them.

Details about primers:
If the beginning of B is TGAATT and the end of A is TGTGTG then the
reverse primer for A should be AATTCA_CACACACA (rev. compl. of B + rev
compl of end of A. Its length should be around 25-30 bp, the longer
the better. I'm using primer3 for these. For graphical explanations
see [2]. Make sure that the overlapping part (here: TGAATT) doesn't
have any overlap with the forward primer of B. The forward primer of B
has to be a little bit 5' of the overlap, otherwise the two primers
will anneal in PCR III.

This is supposed to work for relatively short PCRs, around some kb.
You can fuse several fragments like this, for more than two fragments
and longer fragments, >20kb, see [3] for details. You'll have to use
nested primers then.

Do each PCR I and II  (I prefer a PFU/Taq mix here, 10% of PFU to
reduce the error rate, I'm doing few cycles, ~15-20 cycles), estimate
the yield on a gel, estimate equimolar amounts (roughly). Take some ul
of each PCR tube (say: 1 ul from tube 1 and 4ul from tube 2) and mix
them in a second PCR mix called PCR III, without primers. Run 5 cycles
without primers, add end-to-end primers and do another 20 cycles.

The end-to-end primers can be the original ones from PCR I and II
(forward from I and reverse from II), or better new primers. If you
follow the ideas in [4], the end-to-end primers should have no overlap
with the original primers from PCRs I and II (otherwise you'll see
primer dimers). It's better if they have a higher tm than the original
primers, so you can run PCR III with a higher anneal temp. (65 deg)
and the original primers won't be able to hybridize in this PCR, which
should eliminate any other bands corresponding to A or B products. I
think usually you'll see additional bands after PCR III and have to
gel purify in the end. Nevertheless, [4] got nice bands with high-tm
end-to-end primers

I wonder why people still do combinations of restriction + ligation if
they just want to fuse two fragments of DNA...


[5] PCR-induced (ligase-free) subcloning: a rapid reliable method to
subclone polymerase chain reaction (PCR) products. A R Shuldiner, L A
Scott, and J Roth, NAR 1990
[6] Construction of recombinant DNA by exonuclease recession.
Y S Yang, W J Watson, P W Tucker, and J D Capra Nucleic Acids Res. 1993

---------- Forwarded message ----------
From: Neeru Bhagat <neeru_bhagat from yahoo.com>
Date: 29-May-2007 12:20
Subject: Fwd: optimizing overlap PCR
To: maximilianh from gmail.com

dear Max

I will aslo be doing this kind of PCR in future. Did u get the answers
to your queries and what are those. Please do let me know also.

Thanking You


Note: forwarded message attached.

Luggage? GPS? Comic books?
 Check out fitting gifts for grads at Yahoo! Search.

---------- Forwarded message ----------
From: "sarita mallik" <sarita81 from gmail.com>
To: "Neeru Bhagat" <neeru_bhagat from yahoo.com>
Date: Thu, 24 May 2007 19:01:51 +0530
Subject: Fwd: optimizing overlap PCR

---------- Forwarded message ----------
From: Maximilian Haeussler <maximilianh from gmail.com>
Date: May 18, 2007 10:12 PM
Subject: optimizing overlap PCR
To: methods from magpie.bio.indiana.edu

Hi experts,

I want to do overlap-extension  (fusion) PCR in a medium-scale
project, learning from articles [1], [2], [3] and [4]. I have to
optimize for simplicity (no purification) and highest possible yield
of the final product in ug of DNA (it is meant for electroporation,
I'll need around 10 ug of DNA in the end).
I'm doing nested primers, outer primers at a TM of 70, overlap primers
with TM60, to amplify only the fusion product in the final PCR (as
recommended in [4], I've drawn a schema of this at the end of the
message). This seems to work now though with many different primers.

There is a sufficiently pure product in the end. I don't know,
however, how to optimize my final PCR for maximum yield. (It seems
that I have to use some sort of PFU, I don't know why, perhaps I need
the exonuclease activity to get rid of non-overlapping part C (see

What would you recommend:
- Use rather more polymerase or mix taq/Pfu for higher yield?
- Is it worth buying the highly expensive Phusion/... enzymes or would
you rather buy some cheap polymerase for the final PCR?
- For the extension PCR: NEB is selling a special polymerase, Vent
(exo-) for overlap extensions, do you think this is worth the money?
- Apart from increasing polymerase concentration and primers, is there
anything else that can be done for maximum PCR yield (I cannot find a
good reference for "maximum pcr yield")

schema legend:  -- = template, ~~ = primer, \ = non-overlapping DNA,

                                          \   \   C: tm60
       ----------FRAG1-------     +---------FRAG2-----------------
~~~~~~~             ~~~~-~~~                 ~~           ~~
A tm70:tm60            B  tm60:tm60              D 60       E 70
F tm70

outline of the PCRs:
Frag1 and Frag2 are to be joined into one.
PCR1 on genomic DNA:  primers A+B (anneal at 55)
PCR2 on plasmid:           primers C+D (anneal at 55)

mix 1ul of PCR1 and PCR2 into new tube, then run
10x at 50 deg. without primers, add primers F and E, cycle 30x at 60
deg.  annealing temp. (don't really know if all of this is necessary)
No gel-purification in the end, as there is only one (visible) band.

Any general experiences or references on overlap extension PCR are
also appreciated,


[1] Splicing by overlap extension by PCR using asymmetric
amplification, Gene 1997
[2] Fusion PCR, a One-step variant of the Megaprimer Method of
Mutagenesis, Biotechniques 1998
[3] Construction of long DNA molecules using long PCR-based fusion of
several fragments simultaneously, NAR 2004
[4] Rapid Expression of Functional Genomic Libraries, Journal of
Proteome Res 2006
Maximilian Haeussler,
bioinformatics grad student,
CNRS Gif, France

Maximilian Haeussler,
skype: maximilianhaeussler

Methods mailing list
Methods from net.bio.net

Maximilian Haeussler,
skype: maximilianhaeussler

More information about the Methods mailing list