Overlap extension PCR

Mark nospam at nspam.com
Thu Sep 9 08:18:30 EST 2004

Emily Scott wrote:
> I am hoping users have some suggestions to assist with a problem I am having
> with overlap extension PCR/SOE (splicing by overlap extension) and multimer
> formation.
> I can successfully generate a single 1500 bp product (AB) from my first PCR
> reaction using a primer (A) just 5' of the area of interest and an internal
> primer (B).  I can generate a second, 1000 bp product (CD) from another PCR
> reaction using a primer 3' of the area of interest (D) and an internal
> primer (C).  I gel-purify the 1500 and 100 bp products individually.
> By design the products AB and CD overlap by ~50 bp, so I have been trying to
> generate the overlap extension product AD (2500 bp) by PCR using templates
> AB and CD in the presence of primers A and D.  The overlap product does not
> form (in fact no product is formed).
> I have done significant troubleshooting without any success:
> -primers A and D will amplify a similar sized fragment (2500 bp) that
> originates from a single template (no overlap)
> -all PCR stock reagents work in other reactions
> -denaturing temperature increased, all the way to 99 deg C, to help melt any
> secondary structure
> -annealling temperature decreased, all the way to 45 deg C, to promote
> overlap
> -hot start and withhold primers for 1st three rounds over thermocycling to
> give overlap a chance to form and be extended before amplification
> -addition of DMSO (2% and 5%) to eliminate any secondary structure
> I then went back to check the AB and CD templates generated from PCR
> reactions 1 and 2, which originally ran at 1500 and 1000 bp, respectively,
> and were gel purified.  The AB DNA still had a band at 1500, but now also
> has additional (lighter or less intense) larger bands--at 3000 and 4500 bp.
> To me this suggests multimer formation (concatamers) but I don't know how or
> why they are forming (AB product was stored at -20 deg, no heating/cooling,
> no ligase).  The sequence of the AB PCR product should be:
> 5'-ttt aCC ATG gct aag aaa acg agc tct aag ggg aag ctg ccT CCG Gga  . .  .
>  . . . (~1500 bp) . . .
> aCC ATG agc ttc ctg ccc cgc cat cac cac cat tga taT CCG Gct gag gca gcc-3'
> In capital letters I tried to highlight above two short (5 base pair)
> regions that are identical in the 5' and the 3' ends of the fragment (CCATG
> and TCCGG).  So, multimers could theoretically form, but it seems unlikely
> under any conditions, not to mention while incubating at -20 deg or sitting
> at room temperature.
> And even if multimers are forming, the majority of the AB template is the
> correct 1500 bp size and I don't understand why overlap and extension
> wouldn't occur.  Perhaps the multimer issue a red herring and there is some
> other issue I haven't thought of?
> Thanks,
> Emily

I have done this once, and met with success so have not had the 
need to do loads of trouble shooting. However, I did come across 
an article that may help. You may have tried this already, but if 
not, check out the use of asymmetric PCR in SOE-PCR constructions.

Gene. 1997 Feb 20;186(1):29-35
"Splicing by overlap extension by PCR using asymmetric 
amplification: an improved technique for the generation of hybrid 
proteins of immunological interest."
Warrens AN, Jones MD, Lechler RI.

Good luck.


More information about the Methods mailing list