Supercoiled plasmid pBR322 DNA was purchased from Bio Basic Inc. (Markham Ontario, Canada). Bathocuproine was purchased from Acros Organics (Pittsburgh, PA, USA). The Plasmid Maxi Preparation kit was purchased from Omega Biotek Inc. (Doraville, GA, USA). All the other chemicals were of analytical grade. Pice was dissolved in dehydrated alcohol as a 1 mM stock solution and stored at −20°C in the dark; CuSO₄ was dissolved in ultrapure water as a 1 mM stock solution. Both Pice and CuSO₄ were diluted with 0.01 M phosphate-buffered saline (PBS) prior to use, and the diluted solution was used within 2 h.

DNA strand breakage was measured in terms of conversion of supercoiled pBR322 plasmid DNAs to open circular or linear forms by gel electrophoresis. The reaction mixture (10 μl) contained 100 ng pBR322 DNA, 200 μM CuSO₄ and 200 μM Pice, equal molar antioxidants and metal ion chelating agents in PBS at pH 7.4. After being preincubated for 1 h at 37°C, the reaction mixture was treated with 2 μl loading buffer (6X), and then immediately loaded onto a 1.2% agarose gel containing 1 μg/ml gelview. Horizontal gel electrophoresis was performed in 1X TBE buffer (pH 8.3) and photographed with ultraviolet-visible (UV-visible) transillumination.

UV-visible spectra were measured with a TU-1901 dual beam UV-visible spectrophotometer (Beijing Purkinje General Instrument Co., Ltd., Beijing, China).

A reaction mixture (2 ml) containing 50 μM Pice in PBS (pH 7.4) was prepared and underwent spectral analysis. The spectral tracing was started by adding 1 ml of 50 μM CuSO₄. The spectra were recorded every 5 min after adding CuSO₄ until the variation was invisible.

Varying amounts of Pice and Cu(II) were added to the reaction mixture (2 ml) containing 0.01 M PBS (pH 7.4) and 300 μM bathocuproine, and the final reaction volume was maintained at 3 ml by adding PBS. The bathocuproine-Cu(I) complex was determined by measuring the absorbance at 480 nm following incubation for 10 min at room temperature (33).

Ethylenediaminetetraacetic acid (EDTA), a well-known chelating agent for metal ions, was used to detect the formation of the Pice-Cu(II) complex. The reaction mixture (final volume of 3 ml) contained 50 μM Pice, EDTA and Cu(II) in PBS. The spectra of the reaction mixtures were measured when Cu(II) (50 μM) had reacted for 10 min before and after the addition of EDTA, respectively.

The interaction between Pice and pBR322 DNA was detected at 17°C and 37°C, respectively. The PBS reaction mixtures (3 ml) contained 50 μM Pice and 100, 200 and 400 μl pBR322 DNA (30 ng/μl), respectively. The spectra were measured after incubation for 1 h at 17°C and 37°C, respectively.

The interaction between Pice and pBR322 DNA in the presence of Cu(II) was also detected at 37°C. The PBS reaction mixtures (3 ml) contained 50 μM Pice, Cu(II), and 100, 200 and 400 μl DNA, respectively. The spectra were measured following incubation for 1 h at 37°C.

Fluorescence spectra were recorded on an LS-55 fluorescence spectrophotometer (Perkin-Elmer) in a 1-cm quartz cuvette using an excitation wavelength of 280 nm and an emission wavelength of 300–500 nm. The slit widths for excitation and emission were set at 10 nm. The PBS reaction mixture (final volume of 3 ml) consisted of 50 μM Pice, 0 or 400 μl pBR322 DNA (30 ng/μl) and 50 μM Pice. The spectra were measured following incubation for 1 h at 37°C.

The infrared spectra were recorded as KBr pellets on an FTS 3000 FTIR spectrometer, within a wavelength range of 650–4000 cm^-1, and captured at a spectral range by accumulating 32 scans with a resolution of 4 cm^-1. It should be noted that Pice is easily oxidized during the FTIR spectra measurement, which causes difficulty when measuring the spectra directly, whereas Res is stable and shares a similar structure with Pice. Therefore, Res was used as a control to determine the structure of the product(s) of Pice-Cu(II).

The cleaved supercoiled pBR322 DNA and the open circular and linear DNA were used to assess DNA strand breakage (33). We found that at a concentration of 200 μM, neither Pice nor Cu(II) alone showed the ability to damage DNA (Fig. 1, lanes 2 and 10). However, Cu(II) and Pice worked cooperatively to break the supercoiled DNA (Fig. 1, lane 3).

To clarify the roles of oxygen-derived active species in the DNA strand breakage induced by Pice in the presence of Cu(II), the free radical scavengers and the Cu(I)/Cu(II) chelator were used to prevent DNA strand breakage. Antiscorbic acid (antioxidant), glutathion (GSH, ROS scavenger), mannitol (HO^• scavenger), EDTA [the specific Cu(II) chelator] and bathocuproine (the specific Cu(I) chelator) were all provided to protect pBR322 plasmid DNA against strand breakage induced by Pice in the presence of Cu(II). It was found that antiscorbic acid did not prevent DNA strand breakage (Fig. 1, lane 4), whereas GSH and mannitol completely prevented DNA strand breakage (Fig. 1, lanes 5 and 7), indicating that ROS, particularly HO^•, is important in inducing DNA breakage. Bathocuproine did not effectively protect DNA from breakage (Fig. 1, lane 6), whereas EDTA was able to protect DNA from breakage (Fig. 1, lane 8), although the protection was depressed if EDTA was added after Pice-Cu(II) had reacted with DNA for 15 min (Fig. 1, lane 9). These results suggest that both ROS(HO•) and Cu(II) are critical to DNA damage.

To clarify the mechanism of DNA damage induced by the Pice-Cu(II) system, the UV-visible absorption changes of Pice in the presence of Cu(II) in PBS buffer (pH 7.4) under aerobic conditions were measured (Fig. 2). The peaks of Pice absorption were at 218 and 323 nm (Fig. 2). When Cu(II) was added to Pice, the absorbance at 323 nm disappeared rapidly, accompanied by a red shift at 332 nm, which provided evidence of the formation of the chelate complex Pice-Cu(II). The absorbance at 332 nm decreased with increasing time. A blue shift was also found from 218 to 208 nm with the increasing absorbance, which signaled the formation of new product(s). A new weak peak appeared near 459 nm, due to the formation of the ortho-semiquinone anion. In addition, the ortho-semiquinone radical anion was more easily oxidized to form the final product ortho-quinone (35). Two isosbestic points at 248 and 395 nm and the isosbestic point at 395 nm suggested a direct transition from one form [Pice-Cu(II) complex] to another (ortho-semiquinone anion).

To further understand the mechanism underlying the reaction of the Pice-Cu(II) system, the UV-visible spectrum of the oxidative product(s) of Pice was measured (Fig. 3). In Fig. 3, the spectrum of the oxidative product(s) is identical to that of the reaction product(s) of the Pice-Cu(II) system, indicating that the oxidative product(s) of Pice is one of the reaction product(s) of the Pice-Cu(II) system. Thus, Pice was oxidized and the reducing agent would be Cu(II) in this reaction, and the reaction mechanism was similar to that of other polyphenolic compounds (2).

If the reducing agent was Cu(II), Cu(I) would be formed. To detect Cu(I) in the reaction system, bathocuproine, a selectively trapping agent of Cu(I) was used. The absorbance at 480 nm provided evidence of the formation of the bathocuproine-Cu(I) complex (33). The result of the production of Cu(I) by stoichiometry is shown in Fig. 4. The absorbance at 480 nm was shown to vary with the molar ratio of Pice to Cu(II), indicating the existence of Cu(I) in the Pice-Cu(II) reaction system. No clear maximum was found where the absorption reached a plateau, suggesting that copper ions may be recycled in this reaction (Fig. 4).

As Cu(II) was reduced to Cu(I) by Pice, it is essential to clarify whether the Pice-Cu(II) complex was formed. A well-known chelating agent for metal ions, EDTA, was used to confirm the formation of the Pice-Cu(II) complex (Fig. 5). When Cu(II) was added to Pice, the peak at 323 nm disappeared and the red-shifted peak appeared at 332 nm. Upon addition of EDTA after 10 min, the red-shifted band (332 nm) returned to its initial position (323 nm) with a decrease in absorbance. If EDTA was added to Pice before Cu(II), the spectrum of Pice changed more weakly than it did when EDTA was added after Cu(II), and there was no peak shift. The above phenomena therefore indicate that Pice chelates with Cu(II) as a bidentate ligand, thus facilitating intramolecular electron transfer to form the stable ortho-semiquinone anion (Fig. 2).

The absorption band of electrons is commonly used to study the interactions between micromolecules and DNA. In general, micromolecules possess electron absorption in the UV-visible region, and their electron atmosphere varies in the presence of DNA, which causes the variation of ligand circumstance. These phenomena may be demonstrated using the electron absorption spectrum.

When intercalation occurs between DNA and micromolecules, the absorption intensity of micromolecules degrades, and the absorption peak shows a red shift effect. π electron accumulation may occur between the intercalating ligand and base pairs of DNA, which may cause the conjugation between the π* vacant orbital of micromolecules and the π electron orbital of DNA base pairs. The decrease in the energy level induces π→π* transition energy decrease and generates the red shift effect, and the degree of the red shift reflects the ability of intercalation.

The bathochromic effect occurs following the micromolecule-DNA reaction, and the extent of the bathochromic shift increases with the increasing concentration of DNA, which indicates that the molecular aggregation and hydrogen bond damage cause interactions between micromolecules (36). If the electrostatic interaction and groove binding occur, the weak red shift and hypochromic effect may appear.

Fig. 6 shows the spectra of Pice with different DNA levels at 17°C. Following the addition of pBR322 DNA, the absorbance of Pice at 323 nm decreased with the increasing DNA levels following the weak red shift (3 nm). These changes indicated that the interaction between Pice and DNA was groove binding. The absorption at 218 nm increased, and when the DNA level increased to 400 μl, the blue shift occurred from 218 to 205 nm, suggesting that the katogene in hydrogen bond and molecular aggregation of Pice also existed besides the intercalation (36).

The spectra of Pice with different DNA levels at 37°C are shown in Fig. 7. In contrast to the situation at 17°C, the hypochromic effect and the red shift were evident, indicating that intercalation also occurred between Pice and DNA. Similar to the situation at 218 nm, the hyperchromic effect was further elevated and the absorption peak moved to 208 nm instead of 205 nm with an increase in DNA, indicating that oxidization also occurred in the reaction system. All of these factors suggest that the degree of the interaction between Pice and DNA positively correlated with the temperature.

To further elucidate the mechanism of DNA breakage, the spectra of Pice with different DNA levels at 37°C in the presence of Cu(II) were determined (Fig. 8). The absorption peak at 323 nm disappeared when Cu(II) was present in the reaction system, the blue shift occurred at 218 nm, and the peak moved from 218 to 208 nm accompanied by the increase in absorbance. These results combined with those obtained in the former sections indicate that Pice was oxidized completely in this reaction system. From these results, we hypothesized that the intermediate product mainly caused the damage to DNA.

DNA possessed no endogenous flourescence, although it affected the fluorescence character of the micromolecule through the interactions. In general, the fluorescence intensity was enhanced, and the degree of the enhancement was correlated with the degree of binding. Notably, the rigidity of the micromolecule was enhanced, whereas the micromolecule entered the hydrophobic environment from the hydrophilic environment, which also increased the fluorescence intensity.

The fluorescence spectra of Pice, Pice-DNA and Pice-Cu(II)-DNA at 37°C were measured. As shown in Fig. 9, with the addition of DNA, the fluorescence intensity of Pice increased, mainly due to the hydrophobic protection of DNA. Simultaneously, the architecture of the DNA was damaged.

To determine the structure of Pice, the FTIR spectra of Res, Pice and the reaction product of Res-Cu(II) were measured (Figs. 10 and 11). We found that the medium intensity band emerged in the range 3000-3500 cm^-1, which was attributed to the phenolic hydroxyl group [ν(OH)] of Res; 1630 cm^-1, which was attributed to the stretching vibration of -C=C- outside the aromatic ring; 675 and 806 cm^-1, which was attributed to the out-of-plane bending vibration of =CH in the aromatic ring. The bands of 990–1300 cm^-1 were attributed to the in-plane bending vibration of =C-H in the aromatic ring, and the several sharp bands of 1606, 1585, 1514, 1465 and 1384 cm^-1 were attributed to the in-plane stretching vibration of -C=C- in the aromatic ring. In the produce(s) spectrum of Res, the out-of-plane bending vibration of =CH and the in-plane stretching vibration of -C=C- in the aromatic ring disappeared, and the new band at 1650–1680 cm^-1 appeared, which was attributed to the stretching vibration of quinonyl. These results indicate that the aromatic ring in Res was destroyed when Res reacted with Cu(II). The product was semiquinone, which was the oxidative product of Res. It was also found that the bands of the in-plane bending vibration of =C-H in the aromatic ring varied, which further confirmed the conclusion.

From the FTIR spectra of the reaction product of Res-Cu(II) and Pice-Cu(II) in Fig. 11, two products were shown to have an almost identical spectrum structure, which suggested that the reaction mechanism of Res-Cu(II) and Pice-Cu(II) was identical. When Res was taken into account, oxidized forms of Pice were produced in the presence of Cu(II).