RT-PCR and gels

[Jim Hutchins]

GORDON BETTS (betts at ORION.ETSU.EDU) wrote:
: I believe it is possible to run RT then take 3 ul or so and run
: on a 1% agarose gel to verify you have a product. Correct?

I doubt whether this will work; you would need >1 ug of cDNA to
see a product, and if you are using specific primers this is 
unlikely.

: If not, then how can you check to see if your RT worked before 
: wasting reagents on PCR?

Yes.  The `standard' method is to `spike' the RT reaction, or a 
duplicate reaction consisting of 1/10 of the RT reaction, with
a small amount of 32P-labeled nucleotide (32Palpha-dCTP, e.g.).
Then, from the known molar amounts of each reactant, and making
assumptions about the size of your cDNA, you can calculate a
percent incorporation after separating the reaction from the
cDNA on a spun column (or better, two spun columns).  Most people
use Cerenkov counts to get a guesstimate of the percent incorporated.
In my hands, this number is very, very low.  For example, in a kinase
reaction, we expect 20-50% incorporation, but less that 1% is not
uncommon with an RT reaction.

Anyone else have any comments on this?  What do other labs get as
a percent incorp in a `typical' RT reaction?

Oh, I just remembered, some people use TCA-precipitable counts, which
is cheaper and easier than the spun column unless you already have
them around.

Hope this helps.

--
Jim Hutchins
Assoc Prof Anatomy, Asst Prof Neurology (Research), Univ Mississippi Med Ctr
http://fiona.umsmed.edu/~hutchins/    ***    E-mail:   hutchins at umsmed.edu
``It became necessary to destroy the town in order to save it.''--American
infantry officer firing on Ben Tre, Vietnam, 2/8/68