Matthew L. Brown, Ph.D gemini3 at mindspring.com
Fri Feb 14 13:53:00 EST 1997


Are you doing the PCR on a cDNA library or genomic DNA? If from a
library are you doing the PCR with 1 gene specific primer and another
nonspecific primer like lambda forward or reverse? We routinely amplify
and clone PCR products from cDNA libraries for sequencing. However, like
you I never fully trust the data fully. The reason is that I have seen
several anomolous sequences especially on the nonspecific side of the
product, i.e. lambda arm side. I have seen multiple linker-like regions
which suggest that several adaptors or linkers were ligated to
aparticular cDNA molecule before it was ligated into the lambda arms.
Also I have seen a gene specific PCR product ligated to another
unrelated gene product that could be amplified as a single unit product
from the library. (I suspect this happens a lot but don't know why.)
Anyway, I've seen a lot of anomolies. The trick is to determine what is
a PCR artifact or library artifact and what is real sequence. Our
solution may work in your case. We use some of the cloned PCR products
to make DNA probes for screening (phage hybridization) the library. We
screen the library and select phage clones for sequencing. Many of these
clones will contain the sequence of the probe and additional gene
specific sequence. We use several clones over any given region to
confirm our sequence data (which I think is always necessary). I never
trust a PCR product's sequence, especially near the nonspecific primer,
until I see the same sequence in several library clones. As to whether a
polymerase can jump from one sequence to another, I think that if your
using genomic DNA that polymerases can move from one gene to another
(maybe across hairpins or other secondary structure??). Also if your
using genomic DNA consider the possibility that this "other" gene
sequence is really an intron sequence. As to cDNA library sequencing, as
mentioned above, I have seen anomolous ligation products in libraries.
One in particular stood out because it was a gene specific clone (~1200
bp)that contained the poly-A tail [oligo d(T)primed library] then
immediately after the tail an EcoRI site and linker sequence (identical
to the one used in construction of the library) and then some sequence
(~1000 bp) presumed to be from another gene then another linker sequence
with EcoRI site and then the lambda arm (whooh! that was a long
sentence). Well I hope some of this helps. Feel free to e-mail me if it
helps (or hurts). :-)


Matt Brown, Ph.D.

More information about the Methods mailing list