Design, Synthesis and DNA Interaction Study of New Potential DNA Bis-Intercalators Based on Glucuronic Acid

A series of novel potential DNA bis-intercalators were designed and synthesized, in which two glucuronic acids were linked by ethylenediamine, and the glucuronic acid was coupled with various chromophores, including quinoline, acridine, indole and purine, at the C-1 position. The preliminary binding properties of these compounds to calf thymus DNA (CT-DNA) have been investigated by UV-absorption and fluorescence spectroscopy. The results indicated that all the target compounds can interact with CT-DNA, and the acridine derivative, 3b, showed the highest key selection vector (KSV) value, which suggested that compound 3b binds most strongly to CT-DNA.


Introduction
Numerous biological experiments have suggested that DNA is one of the primary cellular targets for many anticancer agents. Particularly, in cancer cells, DNA can be preferentially damaged, due to the interactions with anticancer agents, therefore inhibition/blockage of cell division causes cell OPEN ACCESS death [1][2][3][4][5]. The DNA interacting molecules are usually bound to DNA non-covalently by three modes: intercalation, groove binding and static electronic interactions. Static electronic interactions refers to molecules that bind with the negatively charged DNA double helix externally through a non-specific interaction. In groove binding, the targeting molecules interact with DNA in the base edges of the major groove or minor groove, which had been discussed by many groups [6][7][8][9][10]. The intercalation is another DNA binding mode that is closely related to the antitumor ability of many anticancer agents [11].
Intercalators are a group of compounds that interact with the DNA double helix in a reversible manner. Some of them are valuable drugs currently used for the treatment of ovarian and breast cancers, as well as acute leukemia, while many others are in different phases of clinical trials. Intercalating agents share some common structural features, such as the presence of planar polyaromatic systems that are inserted between DNA base-pairs with a marked preference for 5'-pyrimidine-purine-3' sequence. In addition, the chromophores are linked to basic moieties that might also play an important role in the affinity and selectivity shown for compounds [12][13][14][15].
Bis-intercalators have two potential intercalating ring systems connected by linkers that can vary in both length and rigidity. In the 1990s, Priebe and co-workers developed the bis-intercalators of daunomycin, which had expanded research ideas [16]. In recent years, the synthesis of bis-intercalators has drawn considerable attention, because, in comparison with mono-intercalators, higher DNA-binding constants, slower dissociation rates and substantial sequence selectivity can be expected by incorporating two or more intercalating units into a polyfunctional ligand [17]. In addition, the interactions with the groove or phosphate groups by the linker may provide enhancement for binding affinity or selectivity [18].
A number of sugar-intercalator conjugates have been studied before for different nucleic acid binding [19][20][21][22][23][24], and monosaccharides have been introduced as biologically relevant scaffolds. The advantage of using saccharides is to acquire the high density of functional groups. Meanwhile, they are available as single enantiomers and contain multiple sites for attachment of recognition groups [25,26]. In this work, we report the design and synthesis of four potential DNA bis-intercalators based on glucuronic acid ( Figure 1). Glucuronic acids, which have a 6-COOH rather than a 6-CH 2 OH, are in glycosides converted to amides via coupling with an amine. The intercalation binding force comes from the π-π interactions and hydrophobic interactions between the aromatic ring of the intercalator and the DNA base, so we chose the planar aromatic molecules, quinoline, acridine, indole and purine, as chromophores to incorporate into these bis-intercalators. The interactions of these bis-intercalators with calf thymus DNA (CT-DNA) were studied.

DNA Binding Properties
To study how efficiently the synthesized compounds interact with CT-DNA, we investigated their DNA binding primary properties by UV absorption and fluorescence emission spectra.

UV Absorption
The measurement of UV absorption of compounds was conducted in the phosphate buffer by using a fixed compound concentration (2 × 10 −5 M) to which increments of the CT-DNA stock solution was added. The solutions were allowed to incubate for 5 min before the absorption spectra were recorded. The UV absorption band of compound 3c and 3d was similar to that of the DNA duplex, around the wavelength of 260 nm; only UV absorption spectra of compound 3a and 3b were measured.
The change in compound 3a and 3b UV absorption spectra with increasing concentration of CT-DNA is shown in Figure 2. Specifically, both compounds showed hypochromicity upon increasing DNA concentration, indicating that both compounds interacted with the DNA helix. However, a weak hypochromic effect is indeed observed, and no red shift is apparent. So, these changes were not significant enough to accurately determine the DNA-binding constants.

Fluorescence Emission
The fluorescence titration experiment has been widely used to characterize the compound-DNA interactions, in which the fluorescence emissions of interacting compounds can be quenched, which results in the decrease of fluorescence intensity [32,33]. On the other hand, in some compounds, the compound-DNA interactions can prevent the compound fluorescence emission from being quenched by polar solvent molecules. Consequently, the fluorescence intensity increases [33]. In this study, the interactions between the compounds and CT-DNA were investigated by fluorescence titration. The results are shown in Figure 3. Specifically, the fluorescence intensity of compound 3a, 3b and 3d decreased when titrated by CT-DNA, being in good agreement with the fluorescence behavior of other intercalators reported in the literature [17]. On the contrary, an increase in the fluorescence intensity of compound 3c was observed. The results indicated that all the compounds can interact with CT-DNA, whereas the binding mode may be different.

EB Competition Assay
The well-established quenching assay based on the displacement of the intercalating dye, ethidium bromide (EB), from CT-DNA was employed to further investigate the interaction mode between the complexes and CT-DNA. EB is a very useful DNA structural probe, which shows a significant increase in fluorescence intensity when intercalating into the base pair of DNA. However, the enhanced fluorescence can be quenched evidently when there is a second complex that can replace the bound EB or break the secondary structure of DNA [33][34][35]. It has been reported that the groove DNA binders can also cause the decrease in EB emission intensities. The effects were, however, only moderate [36].
Before we started the fluorescence test of the compounds with DNA, we tried different excitation wavelengths from 280 to 550 nm and found only one fluoresce for each compound, and each had a fixed wavelength, so we chose the excitation wavelength that is similar with the UV absorbance one. All the compounds do not fluoresce with an excitation at 518 nm.
The EB competition assay results are shown in Figure 4. The fluorescence intensity of DNA-bound EB at 586 nm decreased remarkably with the increase of the compound (3a-3d) concentrations. This decrease in fluorescence intensity may be due to the quenching of some EB molecules that were released from DNA into the solution after being replaced by the compounds.
The similar phenomenon that the fluorescence of DNA-bound EB was quenched as a result of the DNA and compound interactions, a characteristic sign of intercalation [37], has also been reported for the chitosan complex [38].
The fluorescence quenching of DNA-bound EB can be well described by the linear Stern-Volmer equation [39] in which the synthesized compounds were the quenchers: I 0 and I represent the fluorescence intensities in the absence and presence of quencher, respectively; KSV is a linear Stern-Volmer quenching constant; Q is the concentration of quencher. The KSV values were given by the ratio of the slope to intercept. The KSV values for the tested compounds are listed in Table 1. [compound] for the titration of the compound to CT-DNA-EB.
As shown is Table 1, compound 3b had the highest KSV value, which suggested that compound 3b bound most strongly to CT-DNA. The classic intercalators, such as acridine derivatives, seemed to be the most efficient. Further investigations are being currently conducted in our laboratory to obtain an efficient dye group (or proper linker) for the bis-intercalators.

General Methods
All reagents and solvents were purchased commercially and purified in a conventional manner. Thin layer chromatography (TLC) was performed on precoated E. Merck Silica Gel 60 F254 plates. Flash column chromatography was performed on silica gel (200-300 mesh, Qingdao, China). 1 H NMR and 13 C NMR spectra were taken on a Jeol JNM-ECP 600 spectrometer with tetramethylsilane (Me 4 Si) as the internal standard, and chemical shifts are recorded in d-values. Mass spectra were recorded on a Q-TOF Global mass spectrometer.

DNA-Binding Studies
Calf thymus DNA (CT-DNA) was obtained from Sigma (Sigma, St. Louis, MO, USA). All experiments involving CT-DNA were performed in 10 mM phosphate buffered saline (PBS, 100 mM NaCl, pH = 7.0). The per nucleotide DNA concentration was determined by absorption at 260 nm [40] with the molar absorption coefficient of 6600 M −1 cm −1 . A ratio of 260 nm/280 nm around 1.8-1.9 indicated high quality of the CT-DNA sample (free of protein contamination) [41]. Stock DNA solutions (c = 1.67 × 10 −2 M) were stored at 4 °C and for no longer than 4 days before use (1.8 μL solution was added to the cuvette, which makes the concentration of DNA 1 × 10 −5 M, and the added total volume is 10.8 μL when the ratio of compound to DNA is 1:3).
The UV-visible absorption spectra of DNA binding studies were performed on a Shimadzu UV-2550 spectrophotometer (Shimadzu Co., Kyoto, Japan). The fluorescence spectra were measured with a Fp-750w Fluorometer. The absorption and fluorescence titration samples contained small aliquots of DNA solution and the same concentration of the synthesized compound. All the tested compounds were evaluated in 50% DMSO, because they were poorly soluble. The fluorescence spectra of the complexes were recorded by using the excitation wavelength of 300 nm for 3a and 3d, 350 nm for 3b and 290 nm for 3c. The emission wavelength were as follows: 399 nm for 3a, 429 nm for 3b, 324 nm for 3c and 350 nm for 3d. Before measurements, the mixture was mixed well and incubated at room temperature for 5 min.
In the ethidium bromide (EB) fluorescence displacement experiment, 5 μL of the EB Tris solution (1 mM) was added to 1 mL of DNA solution (1 × 10 −4 mol/L, at saturated binding level) [42], followed by a 2 h incubation in the dark. The DNA/compound complex was then titrated into the EB/DNA mixture. Before measurements, the solution was well mixed and incubated at room temperature for 30 min. Fluorescence spectra of EB bound to DNA were obtained at an excitation wavelength of 518 nm and an emission wavelength of 586 nm.

Conclusions
DNA bis-intercalators has drawn considerable attention because of higher DNA-binding constants, slower dissociation rates and substantial sequence selectivity. We designed and synthesized a series of novel potential DNA bis-intercalators, in which two glucuronic acids were linked by ethylenediamine, and the glucuronic acid was coupled with various chromophores, including quinoline, acridine, indole and purine, at the C-1 position. The preliminary binding properties of these compounds to calf thymus DNA (CT-DNA) have been investigated by UV-absorption and fluorescence spectroscopy. The results indicated that all the target compounds can interact with CT-DNA, and the acridine derivative, 3b, showed the highest key selection vector (KSV) value, which suggested that compound 3b binds most strongly to CT-DNA and deserves further development.