cDNA synthesis

Ned Mantei mantei at neuro.biol.ethz.ch
Thu Nov 16 01:59:59 EST 1995


In article <488cba$qk0 at mark.ucdavis.edu>, ez022056 at dale.ucdavis.edu
(Edward Wang) wrote:

> I am running into consistent problems with my cDNA synthesis.  I am 
> reverse transcribing 5 ug of oligo dT selected mRNA (from rat mesangial 
> cells).  First strand looks fine on a gel exposed O/N.  The second 
> strand, on a regular gel stained with etBr shows a very bright band at 
> around 1.5 kb, along with the smear which is expected for a cDNA 
> synthesis rxn of total mRNA.  The band showed up consistently using 
> different batches of reagents and mRNA preps. 

I once spent some months trying to deal with this problem. I found that to
get the product RNA and oligo dT were necessary, but reverse transcriptase
was not. The product was not labelled with 32P-dCTP. The length of the
product varied from one preparation to another. For me the answer was to
use less DNA polymerase than was specified in earlier (ca. pre-1980)
protocols for the 2nd strand: about 25 units Pol I per microgram 1st
strand cDNA (measure as TCA-insoluble cpm incorporated into first strand,
with 32P-dCTP of known specific activity), 2 hr. synthesis. 
There was later a short article in Nucleic Acids Research about this
problem in ca. 1987; one of the authors was Walter Fiers. The conclusion,
which I had also reached, is that the band contains poly(A.T). It probably
occurs when poly(A) from the RNA primes synthesis on an oligodT template.
"Slippage" can occur on such a template, and after awhile a product of
considerable length builds up and is amplfied

Incidentally, in my most recent cDNA synthesis project, after 4 tries in
which one strand or the other wasn't satisfactory, I gave up and bought a
cDNA kit for the first time (from Stratagene). I was impressed--both
strands were long, everything worked the first time. We size-selected the
cDNA on a LGT agarose gel, and starting with 5 micrograms poly(A)+ RNA
could have made 17 million clones with inserts >1kb. I don't think it's
worth spending time titrating components against one another, etc., as
compared to the cost of a kit.

-- 
Ned Mantei
Dept. of Neurobiology, Swiss Federal Institute of Technology
CH-8093 Zurich, Switzerland
mantei at neuro.biol.ethz.ch   Fax: +41-1-633-1046



More information about the Methods mailing list