RT-PCR question

Jim Owens jow at helix.nih.gov
Mon Dec 20 14:55:29 EST 1993


I am attempting to detect chromosomal translocations at the RNA level by
RT-PCR.  I am using RNA from a cultured cell line with a characterized
(i.e., sequenced) translocation to develop the method.  The results are
puzzling in the test system, so I thought I would query the assembled
Magi for their opinions.

The first step is reverse transcription to make cDNA.  There are options
for the reverse transcriptase (RTase) and the primers.  I chose M-MLV
RTase with RNaseH activity with the idea that it was better to destroy
the RNA before running the PCR on the cDNAs.  Rather than using random
primers, I decided to use 23- to 25-mers from the genes on either side of
the chromosomal junction so that one would prime  transcripts from one
strand of the chromosomal DNA and the other would prime transcripts from
the other strand.  

Sure enough, there were cDNAs from both strands.  This was not unexpected
since one of the genes is known to be transcribed with different splice
sites from both strands.  However, when the cDNAs amplified by PCR are
sequenced the same two splice variants appear no matter which primer is
used!  One variant has 37 or 38 bases inserted from out of nowhere at the
junction between the two chromosomes.  The entire sequences of the cDNAs
are not known, but it seems unlikely that the same RNA splices would be
used in both orientations.  Negative controls (without RNA or without
RTase) are as expected, no bands in gel electrophoresis after PCR.

I have started looking at another cell line with a similar translocation.
 As of yet I have no sequences, but I do have PCR fragments that are the
same size no matter which strand is reverse transcribed.  Deja vu all
over again!

Any guesses as to what is going on?  Any suggestions of how to prove I
have some contamination in the PCR?

Thanks in advance,

Jim Owens
___________________________________________________
The Secretary denies any knowledge of my existence.



More information about the Methods mailing list