This is due to binding of the enzyme to the DNA, which will alter the migration rate (and/or cause smearing). This can be remedied by adding an ionic detergent (such as 0.1% SDS) to the sample before loading on the gel. Alternatively, extracting the sample with phenol, or using a `quick spin´ DNA purification column can help. This second cause relates to the tendency of some restriction enzymes to remain bound to their substrate after cleavage. All restriction enzymes have a very tight binding affinity for their recognition sites. And, they all have secondary, and relatively lower affinities for DNA in general, and for the half-site which results from cleavage. The strength of this secondary affinity varies greatly among restriction enzymes. For most restriction enzymes, this secondary affinity is sufficiently low enough to not be observed when digestion products are run on a gel. But, for several restriction enzymes, the secondary binding will cause an alteration of migration rate of the DNA fragments.