Sequencing gel smears

Neil Portwood Neil.Portwood at
Thu Aug 3 10:14:45 EST 2000

Are there any kind souls who can help? I'm performing fragment quantitation
of 32P-labelled multiplex PCR products on denaturing gels. Unfortunately,
in addition to the expected products are present two to three lines of
smear which variously migrate as a smile or a grouch. This symmetry, and
the fact that each smear appears to be specific to one or other of the
expected products, suggests that the smears represent partial renatured
forms of the expected products, their mobility being determined by
differential thermal properties at the center versus the sides of the gel
during the run (they are also not visible on agarose gels). The smears make
quantitation impossible: they sometimes co-migrate with the expected
products in, say, the center of the gel, and confound the analysis by
dividing signal between a specific band and its 'smear' in different
proportions, depending upon where on the gel a sample is loaded, eg in the
center or at the sides. Worse, in lanes where only small amouts of total
PCR product are loaded, only the specific products are visible (ie no
smears), whereas smearing appears where the total amount of loaded product
is more. The PCR (5 microliter volume) is running for only 28 cycles, so
from agarose gel band intensities I'm pretty sure this isn't a straight
overloading problem. I've tried various TBE concentrations between 0.75x to
2x (both gel and buffer), but with little effect. Could this be a buffer
exhaustion problem, in which case could running gradient gels cure the
smears? Just in case anyone feels it's relevant, the TBE and Sangar reagent
are prepared fresh for every run, I do not routinely degass the Sangar and
the acrylamide is an Ultrapure solution from Life Technologies. Any
suggestions are welcome!

Dr. Neil Portwood,
Radiumhemmets Research Laboratory,
Department of Oncology-Pathology,
Karolinska Institute,
CancerCentrum Karolinska R8, 3rd Floor,
Karolinska Hospital,
171 76 Stockholm,

Tel: 46-8-5177 5408/46-8-5177 5460
Fax: 46-8-33 90 31
e-mail: Neil.Portwood at


More information about the Methods mailing list