SOE PCR woes

[Michael Morales]

: In article <Pine.SOL.3.91.960122181815.9403A-100000 at qlink>,
: Chanda Eva R <3erc at QLINK.QUEENSU.CA> wrote:
: >Dear netters:
: >I've tried several times to join 3 PCR fragments together by SOE, or 
: >splicing by overlap extension (a.k.a. megapriming), but without success! 
: >I followed the "standard" protocol (Horton et al,1993. Meth.Enzymol 217: 
: >270-279)i.e. used 1/4 of a gel-purified, Genecleaned, 100uL PCR reaction of 
: >each fragment, 1uM of each of the "outer" primers, 30 cycles of 1 min 
: >@92C, 2 min @60C, 3 min @72C. My fragments overlap by 60 and 33 bp, 
: >respectively, which is more than the protocol recommends--could that be 
: >causing problems? (e.g. by annealing too slowly/inefficiently compared to 
: >shorter overlaps) I'd GREATLY appreciate any practical tips anyone could 
: >give me!! 
: >Please save me from having to go back to old-fashioned splicing by 
: >restriction-ligation subcloning!
: >Thanx,
: >Eva Chanda

:   Drop your annealing temperature from 60C to 50C.  The Horton paper recommends
: an annealing temperature of 50C.  If that doesn't work, then try dropping the
: primer concentration by a factor of 10 as mentioned in another persons follow-
: up to your post.

:   Speaking of woes about SOE PCR, I have been having intermittent luck with it
: by tring to do site directed mutagenesis in a gene.  I produce two fragments,
: call them 5' and 3'.  The 5' fragment is roughly 200bp and the 3' fragment is
: roughly 800bp.  The overlap site is 18bp in length and the mutation site is
: in the 5' primer for the 3' fragment.  The overlap sequence is:

: 	5'-GTAGTCTAGAGACTTACC-3'

: The mutation site is approximately 10bp away from the end of this primer.  I
: have been able to produce both fragments, but I do not seem to be able to pro-
: duce the overlap product.  My PCR protocol calls for:

: 10X Reaction buffer 			: 10ul
: 2.5mM dNTP's	    			:  8ul
: Outside primers (10uM concentration) 	:  1ul/primer
: Taq Polymerase:				:  1ul (5 units/ul)

: For the reaction itself:

: Sterile, ddH2O				:  69.5 ul
: 5' template				:   1 ul (roughly 170ng/ul)
: 3' template				:   1 ul (roughly 250ng/ul)
: MgCl2 (25 mM)				:   8 ul
: reaction master mix (from above)	:  20.5 ul

:   Total volume: 100ul.

:   The thermocylcer is set to run for 35 cycles with the following combinations:

: 	Cycle 1		: 5 min @ 94C
: 	Cycle 2-34	: 1 min @ 94C/2 min @ 50C/3 min @ 72C
: 	Cycle 35	: 10 min @ 72C

:   I have been able to get this to work with one of the mutants that I am trying
: to produce (an Ala mutant), but the rest don't want to work.  I have tried 
: various ideas (different reaction buffers, lower template concentration, dif-
: ferent temperatures, etc).  Is this what I can expect from SOE.  Is it that 
: flaky that some reactions will work and other won't?  If you see anything that
: you might think I am doing wrong or you might think may be worthwhile, let me
: know.  I am at my wits end with this one and my project has stalled because of
: it.

: Thanks in advance,
: Ido

You may be having problems with an untemplated A at the end of your PCR 
product.  I have been using BM's "expand" polymerase with this method & 
it has been working great.  These mixtures of Taq and a proofreading 
polymerase supposedly will take off an untemplated A in these cases.  
Strategene and BRL, among others make a similar product.  You can usually beg
a sample out of them.

Good luck, hope this helps,

Mike Morales
Duke University