Additionally, a conversion of the alkanolamine form to the iminium form was incidentally BEZ235 observed in the presence of a large amount of DNA. We attempted to achieve this conversion but with a large emission shift using a DNA containing an abasic site that serves as the SG binding site. The AP site is produced in living cells by loss of a nucleobase and thus surrounded by an unpaired and two flanking bases, in which the hydrophobic microenvironment would be different from the DNA groove regions. A 0.1 phosphate buffer with pH 8.3 was employed here. At this pH, SG presents mainly in the alkanolamine form and DNA is still stable in the B-form. Nevertheless, this performance has not been realized for the previously used fluorophores. On the other hand, fluorescence quenching even to a greater degree than the corresponding FM-DNA was observed when the flanking sequences were changed to guanines. From the absorption spectra, besides the 336 nm absorption band, the presence of DNA1-Ys also increases the 405 nm and 470 nm absorption bands, as is occurred for the FMDNA. This alteration in the absorption spectra was also observed for the other AP-DNAs. The 405 nm and 470 nm absorption bands result from the SG iminium form. This phenomenon supports that the AP-DNAs as well as the FM-DNAs favor SG conversion from the alkanolamine form to the iminium form. Previously, Maiti et al. also reported that this conversion is possible when the concentration ratio of DNA nucleotide to SG is more than 6. In comparison to with the fluorescence behavior of SG bound to FM-DNA, the converted SG iminium form shows an enhancement in emission when bound to DNA1-Ys and DNA2-Ys and more quenching when bound to DNA3-Ys and DNA4-Ys, meaning that the SG iminium form is preferable to bind to the AP site. As an example in this aspect, we observed that the quenched fluorescence of 1 mM SG by 5 mM FM-DNA at 415 nm was bathochromically recovered at 586 nm only by further addition of 1 mM DNA1-T. No time-dependent spectral evolution was observed after thoroughly mixing DNA1-T and the FMDNA-pretreated SG solution, indicating that the binding of SG to the AP site is very fast. Relative to the AP site-dependent binding evidenced by the enhanced fluorescence responses for DNA1 and DNA2, the greater quenching for DNA3 and DNA4 with guanines and cytosines flanking the AP site does just mean that the SG binding behavior is really related to the presence of the AP site. The quenching should be caused by electron transfer between the excited-state SG bound at the AP site and the nearby guanines because it is widely accepted that guanine is the most easily oxidizable base in DNA. Herein, the possibility of electron transfer was estimated by redox potentials of the involved species. Although the AP site in DNA4-Ys is flanked by cytosines, not guanines, the guanines on the other strand paired with the flanking cytosines should also approach closely to the AP site-bound SG. Thus, the observed quenching for DNA4-Ys can be also explained by the electron transfer mechanism. The electron transfer rate should overwhelm the radiative decay rate for DNA3 and DNA4, which resulted in the observed fluorescence quenching. This electron transfer mechanism could be also employed to explain the lowest fluorescence enhancement that occurred for DNA1-G and DNA2-G in comparison to the corresponding AP-DNAs having the other unpaired bases. In order to evaluate the SG binding mode, we checked the alterations in fluorescence upon adding the electrolyte of NaCl. As shown in Figure 6, addition of NaCl does not seriously affect the emission of SG bound to DNA1-Ys, whereas NaCl induces a concentration-dependent increase in fluorescence for the FMDNA, indicating release of the bound SG from the FM-DNA upon.