248 Gray and Lu
Whether the modifying treatment itself introduces strand breaks is
readily ascertained by electrophoresis of the treated form I ° DNA and
nontreated controls in agarose gels containing a low concentration of
ethidium bromide (e.g., 11). Ethidium bromide at 1 ~g/mL should also
be present in nuclease reaction mixtures to ensure that changes in the
ionic environment and/or temperature between the conditions of the
nuclease reaction and the topoisomerase treatment (35) do not result
in even slight negative supercoiling. High enzyme concentrations, on
the order of 100 U/mL, should initially be used to provide a stringent
test, because the rate of attack in response to various types of lesions
varies widely.
The rate of disappearance of form I ° DNA, as by agarose gel electro-
phoresis of aliquots containing equal volumes of the reaction mixture,
is the parameter to be assayed, since nuclease at such concentrations
will rapidly degrade, via the duplex exonuclease activity, the linear
duplexes produced after the introduction of a strand break, and cleav-
age in the second strand. In fact, this reaction has been so rapid that it
led to a novel coupled assay in one study in which the loss of fluores-
cence of EtdBr owing to the complete digestion of a DNA of 10,000
bp was used to monitor the rate of cleavage in response to the presence
of apurinic sites; this digestion was fast compared to the rate of intro-
duction of the initial scission (14). In the only study where direct
comparisons were made, the F nuclease was more efficient in cleaving
in response to a covalent lesion than the S species (14).
The exonuclease activity unfortunately destroys information as to
the site of the initial cleavage in reactions converting an appreciable
percentage of the DNA. For a type of distortion that readily elicits
cleavage, that resulting from negative supercoiling, it has proven
possible to localize sites of BAL nuclease cleavage in experiments in
which the overall extent of cleavage was very limited and the DNA
was labeled with 32p after cleavage with restriction nucleases so that
the small fraction undergoing endonucleolytic attack could be exam-
ined in autoradiograms of agarose gels (36). The interesting possibil-
ity of using nuclease that has been treated with protease so that the bulk
of the duplex exonuclease activity is eliminated in order to localize
sites of attack was noted in an earlier section. This will require further
experimentation to optimize the conditions for such protease treatment.