Stratagene Robocycler, optimizing long PCR

John Ladasky ladasky at leland.Stanford.EDU
Thu Oct 17 05:17:33 EST 1996


Greetings, everyone,

	I have some PCR amplifications I'm working on in the 5 Kb - 7 Kb
range.  They are proving to be difficult to optimize by temperature alone,
so I am now going to look at temperature in conjunction with cosolvents,
magnesium concentration, etc.  For enzymes, I'm using Perkin Elmer Taq
polymerase and Stratagene's Taq Extender (Pfu polymerase).  A neighboring
lab has generously offered to let me use their Robocycler 40, which has a
gradient temerature block.  In theory, this should speed up my work tre-
mendously.  Unfortunately, the folks who own the Robocycler have *lost the
manual*, and I'm working on the basis of rules of thumb.

	I have started conservatively, and have met with disaster already.
One of my lab mates regularly performs a 3.1 Kb PCR that only works with
Taq Extender, not with Taq alone.  I have successfully duplicated this 
reaction on our lab's workhorse thermal cycler, a Perkin Elmer 9600.  Then
I took the exact same reaction mixtures over to the Robocycler.  I got 
smears instead of bands.  In my experience, this can happen in PCR's when
DNA is broken by lengthy incubation a high temperatures. giving non-speci-
fic ends that can anneal and extend and compete with the desired reactions.

	I think that the problem may be with the cycling program.  Anyone
who works with the Robocycler should know that one does not enter the time
at which a sample sits at a temperature -- rather, one enters the total 
time that a sample sits on the block, including ramp time.  I was told that
the rule of thumb is to assume a ramp rate of one second per degree Celsius.
Thus if I want the sample to incubate for 30 seconds at 95 degrees following
a 72-degree extension, I would enter a time of 53 seconds (23 seconds to 
ramp plus the 30 second "hold").

	This formula seems too simplistic to me.  Could I be overexposing my
DNA to destructive high temperatures?  Is there a better rule of thumb for 
calculating the time on the block?  Do sample volumes matter?  Are there any
other problems I may have overlooked?  Please reply if you have any ideas.
Thanks a lot!

-- 
Unique ID : Ladasky, John Joseph Jr.
Title     : BA Biochemistry, U.C. Berkeley, 1989  (Ph.D. perhaps 1998???)
Location  : Stanford University, Dept. of Structural Biology, Fairchild D-105
Keywords  : immunology, music, running, Green



More information about the Methods mailing list