Is is possible that your probe contains fragments from its vector? In
several instances, we found that our probes contained fragments from the
bluescript vector that we used to maintain the probe fragment.
These fragments hybridized quite strongly with the vector in some of our
experiments. Most of the commonly used vectors today are derived from
pBR322, so many vectors could cause this type of problem.
Separating probe and vector fragments was not always easy. Sometimes
we had to run the isolated fragment through a second agarose gel to make
sure that it contained no vector fragments. It may also be possible that
some of your reagents contain a contaminating nuclease at low concentration
that produce small fragments of the vector that are not visible in an
agarose gel (although this may be unlikely).
I hope this information will be helpful.
(21848hjk at msu.edu)